A EPA
EPA Document# EPA-740-R-24-012
September 2024
United States Office of Chemical Safety and
Environmental Protection Agency Pollution Prevention
Risk Evaluation for
Tris(2-chloroethyl) Phosphate
(TCEP)
CASRN 115-96-8
September 2024
-------
TABLE OF CONTENTS
ACKNOWLEDGEMENTS 20
EXECUTIVE SUMMARY 21
1 INTRODUCTION 25
1.1 Scope of the Risk Evaluation 25
1.1.1 Life Cycle and Production Volume 25
1.1.2 COUs Included in the Risk Evaluation 29
1.1.2.1 Conceptual Models 31
1.1.3 Populations Assessed 36
1.1.3.1 Potentially Exposed or Susceptible Subpopulations 36
1.2 Systematic Review 37
1.3 Organization of the Risk Evaluation 39
2 CHEMISTRY AND FATE AND TRANSPORT 40
2.1 Physical and Chemical Properties 40
2.2 Environmental Fate and Transport 42
2.2.1 Fate and Transport Approach and Methodology 42
2.2.2 Summary of Fate and Transport Assessment 44
2.2.3 Weight of Scientific Evidence Conclusions for Fate and Transport 47
2.2.3.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the Fate and
Transport Assessment 47
3 RELEASES AND CONCENTRATIONS IN THE ENVIRONMENT 48
3.1 Approach and Methodology 48
3.1.1 Industrial and Commercial 48
3.2 Environmental Releases 52
3.2.1 Industrial and Commercial 52
3.2.1.1 Summary of Daily Environmental Release Estimates 53
3.2.2 Consumer Releases 60
3.2.3 Weight of Scientific Evidence Conclusions for Environmental Releases from Industrial,
Commercial, and Consumer Sources 60
3.2.3.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the
Environmental Release Assessment 60
3.3 Concentrations in the Environment 62
3.3.1 Ambient Air Pathway 63
3.3.1.1 Measured Concentrations in Ambient Air 64
3.3.1.2 EPA Modeled Concentrations in Ambient Air and Air Deposition
(IIO AC/AERMOD) 64
3.3.1.2.1 Partitioning between Gaseous Phase and Particulate Phase 66
3.3.2 Water Pathway 67
3.3.2.1 Geospatial Analyses of Environmental Releases 67
3.3.2.1.1 Geospatial Analysis for Tribal Exposures 68
3.3.2.2 Measured Concentrations in Surface Water 70
3.3.2.3 Measured Concentrations in Precipitation 71
3.3.2.4 Measured Concentrations in Surface Water Databases 72
3.3.2.5 EPA Modeled Surface Water Concentrations (E-FAST 2014, VVWM-PSC) 74
3.3.2.6 EPA Modeled Surface Water Concentrations via Air Deposition (AERMOD) 77
Page 2 of 638
-------
3.3.2.7 Measured Concentrations in Wastewater 78
3.3.2.8 Measured Concentrations in Sediment 79
3.3.2.9 EPA Modeled Sediment Concentrations (VVWM-PSC) 80
3.3.2.10 EPA Modeled Sediment Concentrations via Air Deposition (AERMOD) 82
3.3.3 Land Pathway 82
3.3.3.1 Measured Concentrations in Soil 83
3.3.3.2 EPA Modeled Soil Concentrations via Air Deposition (AERMOD) 83
3.3.3.3 Measured Concentrations in Biosolids 84
3.3.3.4 EPA Calculated Soil Concentrations via Biosolids 84
3.3.3.5 EPA Modeled Soil Concentrations via BST 85
3.3.3.6 Measured Concentrations in Groundwater 85
3.3.3.7 Measured Concentrations in Groundwater Databases 86
3.3.3.8 EPA Modeled Groundwater Concentrations via Leaching (DRAS) 88
3.4 Concentrations of TCEP in the Indoor Environment 90
3.4.1 Indoor Air Pathway 91
3.4.1.1 Measured Concentrations in Indoor Air 91
3.4.1.2 Measured Concentrations in Personal Air 93
3.4.1.3 EPA Modeled Indoor Concentrations as a Ratio of Ambient Air 93
3.4.1.4 Reported Modeled Concentrations in Indoor Air 94
3.4.2 Indoor Dust Pathway 95
3.4.2.1 Measured Concentrations in Indoor Dust 95
3.4.2.2 Reported Modeled Concentrations in Indoor Dust 97
4 ENVIRONMENTAL RISK ASSESSMENT 98
4.1 Environmental Exposures 99
4.1.1 Approach and Methodology 99
4.1.2 Exposures to Aquatic Species 100
4.1.2.1 Measured Concentrations in Aquatic Species 100
4.1.2.2 Calculated Concentrations in Aquatic Species 101
4.1.2.3 Modeled Concentrations in the Aquatic Environment 102
4.1.3 Exposures to Terrestrial Species 103
4.1.3.1 Measured Concentrations in Terrestrial Species 103
4.1.3.2 Modeled Concentration in the Terrestrial Environment 104
4.1.4 Trophic Transfer Exposure 104
4.1.5 Weight of Scientific Evidence Conclusions for Environmental Exposures 106
4.1.5.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the
Environmental Exposure Assessment 106
4.2 Environmental Hazards 109
4.2.1 Approach and Methodology 109
4.2.2 Aquatic Species Hazard 110
4.2.3 Terrestrial Species Hazard 115
4.2.4 Environmental Hazard Thresholds 118
4.2.4.1 Aquatic Species COCs Using Empirical and SSD Data 118
4.2.4.2 Terrestrial Species Hazard Values 119
4.2.5 Summary of Environmental Hazard Assessment 120
4.2.6 Weight of Scientific Evidence Conclusions for Environmental Hazards 122
4.2.6.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the
Environmental Hazard Assessment 122
4.3 Environmental Risk Characterization 126
Page 3 of 638
-------
4.3.1 Risk Characterization Approach 128
4.3.1.1 Risk Characterization Approach for Trophic Transfer 131
4.3.2 Risk Characterization for Aquatic Receptors 134
4.3.3 Risk Characterization for Terrestrial Receptors 139
4.3.4 Risk Characterization Based on Trophic Transfer in the Environment 140
4.3.5 Connections and Relevant Pathways from Exposure Media to Receptors 144
4.3.5.1 Aquati c Receptors 144
4.3.5.2 Terrestrial Receptors 144
4.3.6 Summary of Environmental Risk Characterization 146
4.3.6.1 COUs/OESs with Quantitative Risk Estimates 146
4.3.6.2 COUs/OESs without Quantitative Risk Estimates 151
4.3.7 Overall Confidence and Remaining Uncertainties Confidence in Environmental Risk
Characterization 155
4.3.7.1 Trophic Transfer Confidence 155
4.3.7.2 Risk Characterization Confidence 159
5 HUMAN HEALTH RISK ASSESSMENT 161
5.1 Human Exposures 162
5.1.1 Occupational Exposures 163
5.1.1.1 Approach and Methodol ogy 163
5.1.1.2 Summary of Inhalation Exposure Assessment 168
5.1.1.3 Summary of Dermal Exposure Assessment 173
5.1.1.4 Weight of Scientific Evidence Conclusions for Occupational Exposure 174
5.1.1.4.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the
Occupational Exposure Assessment 175
5.1.2 Consumer Exposures 177
5.1.2.1 Approach and Methodol ogy 177
5.1.2.2 Consumer COUs and Exposure Scenarios 178
5.1.2.2.1 Consumer Exposure Routes Evaluated 183
5.1.2.2.2 Inhalation Exposure Assessment 187
5.1.2.2.3 Dermal Exposure Assessment 189
5.1.2.2.4 Oral Exposure Assessment 191
5.1.2.2.5 Qualitative Exposure Assessment 193
5.1.2.3 Summary of Consumer Exposure Assessment 196
5.1.2.4 Weight of Scientific Evidence Confidence for Consumer Exposure 199
5.1.2.4.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the
Consumer Exposure Assessment 202
5.1.3 General Population Exposures 207
5.1.3.1 Approach and Methodology 208
5.1.3.1.1 General Population Exposure Scenarios 212
5.1.3.2 Summary of Inhalation Exposure Assessment 213
5.1.3.3 Summary of Dermal Exposure Assessment 215
5.1.3.3.1 Incidental Dermal from Swimming 215
5.1.3.3.2 Incidental Dermal Intake from Soil 216
5.1.3.4 Summary of Oral Exposures Assessment 217
5.1.3.4.1 Drinking Water Exposure 217
5.1.3.4.2 Fish Ingestion Exposure 222
5.1.3.4.3 Subsistence Fish Ingestion Exposure 226
5.1.3.4.4 Tribal Fish Ingestion Exposure 226
Page 4 of 638
-------
5.1.3.4.5 Incidental Oral Ingestion from Soil 228
5.1.3.4.6 Incidental Oral Ingestion from Swimming 229
5.1.3.4.7 Human Milk Exposure 231
5.1.3.4.8 Dietary Exposure (Non-TSCA) 233
5.1.3.5 Exposure Reconstruction Using Human Biomonitoring Data and Reverse Dosimetry 235
5.1.3.6 Summary of General Population Exposure Assessment 239
5.1.3.6.1 General Population Exposure Results 239
5.1.3.7 Weight of Scientific Evidence Conclusions for General Population Exposure 243
5.1.3.7.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the
General Population Exposure Assessment 244
5.1.3.7.2 Strengths, Limitations, and Key Sources of Uncertainty for the Human Milk
Pathway 247
5.1.4 Aggregate Exposure Scenarios 249
5.1.5 Sentinel Exposures 251
5.2 Human Health Hazard 252
5.2.1 Approach and Methodology 252
5.2.2 Toxicokinetics Summary 254
5.2.3 Non-cancer Hazard Identification and Evidence Integration 256
5.2.3.1 Key Human Health Hazard Outcomes 256
5.2.3.1.1 Neurotoxicity 256
5.2.3.1.2 Reproductive Toxicity 260
5.2.3.1.3 Kidney Toxicity 264
5.2.3.2 Other Human Health Hazard Outcomes 266
5.2.3.2.1 Skin and Eye Irritation 266
5.2.3.2.2 Mortality 267
5.2.3.2.3 Liver 268
5.2.3.2.4 Immune/Hematological 270
5.2.3.2.5 Thyroid 272
5.2.3.2.6 Endocrine (Other) 273
5.2.3.2.7 Lung/Respiratory 273
5.2.3.2.8 Body Weight 274
5.2.3.2.9 Developmental Toxicity 275
5.2.4 Cancer Hazard Identification, MOA Analysis, and Evidence Integration 282
5.2.4.1 Human Evidence 282
5.2.4.2 Animal Evidence 282
5.2.4.3 MOA Summary 284
5.2.4.4 Evidence Integration Summary 286
5.2.5 Dose-Response Assessment 287
5.2.5.1 Selection of Studies and Endpoints for Non-cancer Toxicity 288
5.2.5.1.1 Non-cancer Points of Departure for Acute Exposure 289
5.2.5.1.2 Non-cancer Points of Departure for Intermediate and Chronic Exposures 292
5.2.5.1.3 Uncertainty Factors Used for Non-cancer Endpoints 300
5.2.5.2 Selection of Studies and Endpoint Derivation for Carcinogenic Dose-Response
Assessment 300
5.2.6 Weight of Scientific Evidence Conclusions for Human Health Hazard 301
5.2.6.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the Hazard
Identification and Selection of PODs for Human Health Hazard Assessment 302
5.2.6.1.1 Acute Non-cancer 302
5.2.6.1.2 Intermediate and Chronic Non-cancer 303
Page 5 of 638
-------
5.2.6.1.3 Cancer 305
5.2.6.2 Human Health Hazard Confidence Summary 306
5.2.7 Toxicity Values Used to Estimate Risks from TCEP Exposure 306
5.2.8 Hazard Considerations for Aggregate Exposure 318
5.2.9 Genotoxicity Hazard Identification and Evidence Integration 318
5.3 Human Health Risk Characterization 320
5.3.1 Risk Characterization Approach 320
5.3.1.1 Estimation of Non-cancer Risks 321
5.3.1.2 Estimation of Cancer Risks 322
5.3.2 Summary of Human Health Risk Characterization 322
5.3.2.1 Summary of Risk Estimates for Workers 322
5.3.2.1.1 COUs/OESs with Quantitative Risk Estimates 323
5.3.2.1.2 COUs/OESs Without Quantitative Risk Estimates 328
5.3.2.2 Summary of Risk Estimates for Consumers 330
5.3.2.2.1 COUs with Quantitative Risk Estimates 330
5.3.2.2.2 COUs Without Quantitative Risk Estimates 335
5.3.2.3 Summary of Risk Estimates for the General Population 336
5.3.2.3.1 COUs with Quantitative Risk Estimates 336
5.3.2.3.2 COUs Without Quantitative Risk Estimates 350
5.3.2.4 Summary of Risk Estimates for Infants from Human Milk 353
5.3.3 Risk Characterization for PESS 355
5.3.4 Risk Characterization for Aggregate and Sentinel Exposures 365
5.3.5 Overall Confidence and Remaining Uncertainties in Human Health Risk
Characterization 368
5.3.5.1 Occupational Risk Estimates 368
5.3.5.2 Consumer Risk Estimates 370
5.3.5.3 General Population Risk Estimates 374
5.3.5.4 Hazard Values 378
6 UNREASONABLE RISK DETERMINATION 384
6.1 Unreasonable Risk to Human Health 386
6.1.1 Populations and Exposures EPA Assessed to Determine Unreasonable Risk to Human
Health 386
6.1.2 Summary of Human Health Effects 386
6.1.3 Basis for EPA's Determination of Unreasonable Risk to Human Health 387
6.1.4 Workers 388
6.1.5 Consumers 389
6.1.6 General Population 391
6.2 Unreasonable Risk to the Environment 394
6.2.1 Populations and Exposures EPA Assessed to Determine Unreasonable Risk to the
Environment 394
6.2.2 Summary of Environmental Effects 394
6.2.3 Basis for EPA's Determination of Unreasonable Risk of Injury to the Environment 395
6.3 Additional Information Regarding the Basis for the Unreasonable Risk Determination 396
REFERENCES 402
APPENDICES 438
Appendix A ABBREVIATIONS, ACRONYMS, AND GLOSSARY OF SELECT TERMS 438
Page 6 of 638
-------
A. I Key Abbreviations and Acronyms 438
A.2 Glossary of Select Terms 441
Appendix B REGULATORY AND ASSESSMENT HISTORY 443
B. 1 Federal Laws and Regulations 443
B,2 State Laws and Regulations 444
B.3 International Laws and Regulations 445
B.4 Assessment History 446
Appendix C LIST OF SUPPLEMENTAL DOCUMENTS 448
Appendix D CONDITIONS OF USE DESCRIPTIONS 452
D. 1 Manufacturing (Import) 452
D.2 Processing - Incorporation into Formulation, Mixture, or Reaction Product - Paint and
Coating Manufacturing 452
D.3 Processing - Incorporation into Formulation, Mixture, or Reaction Product - Polymers used
in Aerospace Equipment and Products 452
D.4 Processing - Incorporation into Article - Aerospace Equipment and Products and
Automotive Articles and Replacement Parts Containing TCEP 453
D,5 Processing - Recycling 453
D.6 Distribution in Commerce 453
D.7 Industrial Use - Other Use - Aerospace Equipment and Products and Automotive Articles
and Replacement Parts Containing TCEP 453
D.8 Industrial Use - Paints and Coatings 454
D.9 Commercial Use - Other Use - Aerospace Equipment and Products and Automotive Articles
and Replacement Parts Containing TCEP 454
D.10 Commercial Use - Paints and Coatings 455
D.l I Commercial Use - Laboratory Chemicals 455
D.12 Commercial Use - Furnishing, Cleaning, Treatment/Care Products - Fabric and Textile
Products 455
D.l 3 Commercial Use - Furnishing, Cleaning, Treatment/Care Products - Foam Seating and
Bedding Products 456
D.l4 Commercial Use - Construction, Paint, Electrical, and Metal Products -
Building/Construction Materials - Insulation 456
D.l5 Commercial Use - Construction, Paint, Electrical, and Metal Products -
Building/Construction Materials - Wood and Engineered Wood Products - Wood Resin
Composites 456
D. 16 Consumer Use - Paints and Coatings, Including Those Found on Automotive Articles and
Replacement Parts 457
D. 17 Consumer Use - Furnishing, Cleaning, Treatment/Care Products - Fabric and Textile
Products 457
D, 18 Consumer Use - Furnishing, Cleaning, Treatment/Care Products - Foam Seating and
Bedding Products 457
D. l 9 Consumer Use - Construction, Paint, Electrical, and Metal Products - Building/Construction
Materials - Insulation 458
D.20 Consumer Use - Construction, Paint, Electrical, and Metal Products - Building/Construction
Materials - Wood and Engineered Wood Products - Wood Resin Composites 458
D.21 Disposal 458
Page 7 of 638
-------
Appendix E DETAILED EVALUATION OF POTENTIALLY EXPOSED OR
SUSCEPTIBLE SUBPOPULATIONS 459
E.l PESS Based on Greater Exposure 459
E.2 PESS Based on Greater Susceptibility 462
Appendix F PHYSICAL AND CHEMICAL PROPERTIES AND FATE AND TRANSPORT
DETAILS 469
F.l Physical and Chemical Properties Evidence Integration 469
F.l.l Physical Form 470
F.l.2 Vapor Pressure 471
F.l.3 Vapor Density 471
F.l.4 Water Solubility 471
F. 1.5 Logarithmic Octanol:Air Partition Coefficient (log Koa) 471
F.l.6 Henry's Law Constant 471
F.l.7 Flash Point 471
F.l.8 Autoflammability 472
F.2 Fate and Transport 472
F.2.1 Approach and Methodology 472
F.2.1.1 EPI Suite™Model Inputs 472
F.2.1.2 Fugacity Modeling 473
F.2.1.3 OECD Pov and LRTP Screening Tool 474
F. 2.1.4 Evi dence Integrati on 475
F.2.2 Air and Atmosphere 475
F.2.3 Aquatic Environments 477
F.2.3.1 Surface Water 477
F.2.3.2 Sediments 479
F.2.3.3 Key Sources of Uncertainty 480
F.2.4 Terrestrial Environments 480
F.2.4.1 Soil 480
F.2.4.2 Groundwater 482
F.2.4.3 Landfills 482
F.2.4.4 Biosolids 483
F.2.4.5 Key Sources of Uncertainty 483
F.2.5 Persistence Potential 483
F.2.5.1 Destruction and Removal Efficiency 483
F.2.5.2 Removal in Wastewater 484
F.2.5.3 Removal in Drinking Water Treatment 485
F.2.6 Bioaccumulation Potential 486
F.2.6.1 Key Sources of Uncertainty 488
Appendix G ENVIRONMENTAL HAZARD DETAILS 490
G. 1 Approach and Methodology 490
G.2 Hazard Identification 490
G.2.1 Aquatic Hazard Data 490
G.2.1.1 Web-Based Interspecies Correlation Estimation (Web-ICE) 490
G.2.1.2 Invertebrate and Vertebrate Web-ICE 490
G.2.1.3 Algal Web-ICE 492
G.2.1.4 Species Sensitivity Distribution (SSD) 496
G.2.1.5 Vertebrate and Invertebrate SSD 496
Page 8 of 638
-------
G.2.1.6 Algal SSD 496
G.2.2 Terrestrial Hazard Data 505
G.2.3 Evidence Integration 506
G.2.3.1 Weight of Scientific Evidence 507
Appendix H ENVIRONMENTAL RISK DETAILS 511
H. 1 Risk Estimation for Aquatic Organisms 511
H.2 Risk Estimation for Terrestrial Organisms 522
H.3 Trophic Transfer Analysis Results 523
Appendix I GENERAL POPULATION EXPOSURE DETAILS 526
I.1 Exposure Factors 526
1.2 Water Pathway 527
1.2.1 Surface Water and Groundwater Monitoring Database Retrieval and Processing 527
1.2.1.1 Water Plots and Figures Generated in R 527
1.2.2 Methodology for Obtaining New Flow Data (2015 to 2020) 529
1.2.3 E-FAST: Predicted Flowing Surface Water Concentrations (First Tier Modeling) 529
1.2.3.1 E-FAST 2014 Exposure Activity Parameters 533
1.2.4 VVWM-PSC: Predicted Flowing Surface Water Concentrations (Second Tier Modeling) 534
1.3 Ambient Air Pathway 534
1.3.1 Modeling Approach for Estimating Concentrations in Ambient Air 535
1.3.2 Ambient Air: Screening Methodology 535
1.3.3 Ambient Air: AERMOD Methodology 538
1.4 Fish Ingestion Pathway 547
1.4.1 Exposure Estimates 547
1.4.2 Risk Estimates 550
1.5 Human Milk Pathway 553
1.5.1 Verner Model 554
1.5.2 Milk Ingestion Rates by Age 558
1.5.3 Modeled TCEP Concentrations in Milk 558
1.5.4 Infant Exposure Estimate 559
1.5.5 Infant Risk Estimates 568
1.5.6 Sensitivity Analysis 576
1.6 Landfill Analysis Using DRAS 580
Appendix J CONSUMER EXPOSURE DETAILS 583
J.l Approach and Methodology 583
J. 1.1 Consumer Exposure Model (CEM) 583
J. 1.2 Inputs 585
J. 1.2.1 CEM and Sensitivity Analysis 585
J. 1.3 Results 585
J. 1.3.1 Raw Consumer Modeling Results 585
J.1.3.1 CEM 3.2 User Guide and Appendices 587
Appendix K HUMAN HEALTH HAZARD DETAILS 588
K. 1 Toxicokinetics and PBPK Models 588
K.l.l Absorption 588
K.1.2 Distribution 589
K.1.3 Metabolism 589
K.1.4 Elimination 590
Page 9 of 638
-------
K. 1.5 PBPK Modeling Approach 590
K.2 Detailed Mode of Action Information 590
K.2.1 Mutagenicity 591
K.2.2 Other Modes of Action 592
K.2.3 Mode of Action Conclusions 593
K.3 Dose-Response Derivation 593
K.3.1 Adjustments for All PODs (Non-cancer and Cancer) 593
K.3.2 Non-cancer Dose-Response Modeling 594
K.3.2.1 Calculating Daily Oral Human Equivalent Doses (HEDs) 594
K.3.2.2 Use of Oral HED as Dermal HED 595
K.3.2.3 Extrapolating to Inhalation Human Equivalent Concentrations (HECs) 595
K.3.2.4 TCEP Non-cancer HED and HEC Calculations for Acute Exposures 596
K.3.2.5 TCEP Non-cancer HED and HEC Calculations for Intermediate and Chronic
Exposures 596
K.3.3 Cancer Dose-Response Modeling 596
K.3.3.1 Calculating Daily Oral CSFs 597
K.3.3.2 Use of Oral CSF as Dermal CSF 597
K.3.3.3 Extrapolating to Inhalation Unit Risks (IURs) 597
K.3.3.4 CSF and IUR Calculations for Lifetime Exposures 598
Appendix L EVIDENCE INTEGRATION FOR HUMAN HEALTH OUTCOMES 599
L.l Evidence Integration Tables for Major Human Health Hazard Outcomes 600
L.2 Evidence Integration Statements for Health Outcomes with Limited Data 623
Appendix M GENOTOXICITY DATA SUMMARY 625
M. 1.1 Chromosomal Aberrations 625
M. 1.1.1 In Vivo Data 625
M. 1.1.2 In Vitro Data 626
M. 1.2 Gene Mutations 626
M. 1.2.1 In Vitro Studies 626
M. 1.3 Other Genotoxicity Assays 627
Appendix N EXPOSURE RESPONSE ARRAY FOR HUMAN HEALTH HAZARDS 633
Appendix O OCCUPATIONAL EXPOSURE VALUE DERIVATION AND ANALYTICAL
METHODS USED 634
O. I Occupational Exposure Value Calculations 634
0.2 Summary of Air Sampling Analytical Methods Identified 637
LIST OF TABLES
Table 1-1. Conditions of Uses in the Risk Evaluation for TCEP 30
Table 2-1. Physical and Chemical Properties of TCEP 40
Table 2-2. Environmental Fate Properties of TCEP 43
Table 3-1. Crosswalk of COUs to OESs Assessed 49
Table 3-2. Summary of EPA's Daily Release Estimates for Each OES and EPA's Overall Confidence in
these Estimates for 2,500 lb Production Volume 54
Table 3-3. Summary of EPA's Release Estimates for Each COU/OES and EPA's Overall Confidence in
these Estimates 57
Page 10 of 638
-------
Table 3-4. Excerpt of Ambient Air Modeled Concentrations and Deposition for the Use of Paints and
Coatings - Spray Application OES, 2,500 lb Production Volume, 95th Percentile Release
Estimate, Suburban Forest Land Category Scenario 66
Table 3-5. Summary of Modeled Surface Water Concentrations for the 2,500 lb, High-End Release
Estimates 76
Table 3-6. Summary of Modeled Benthic Pore Water and Sediment Concentrations for the 2,500 lb
Production Volume, High Estimate Releases 81
Table 3-7. BST Modeled Soil Concentrations for Incorporation into Paints and Coatings 85
Table 3-8. Potential Groundwater Concentrations (|ig/L) of TCEP Found in Wells within 1 Mile of a
Disposal Facility Determined Using the DRAS Model 89
Table 4-1. TCEP Fish Concentrations Calculated from VVWM-PSC Modeled Industrial and
Commercial TCEP Releases 102
Table 4-2. Aquatic Vertebrate Environmental Hazard Studies for TCEP 113
Table 4-3. Aquatic Invertebrate Environmental Hazard Studies for TCEP 114
Table 4-4 Aquatic Plant Environmental Hazard Studies for TCEP 115
Table 4-5. Terrestrial Organisms Environmental Hazard Studies Used for TCEP 117
Table 4-6. Environmental Hazard Thresholds for Aquatic Environmental Toxicity 121
Table 4-7. Environmental Hazard Thresholds for Terrestrial Environmental Toxicity 121
Table 4-8. TCEP Evidence Table Summarizing the Overall Confidence Derived from Hazard
Thresholds 125
Table 4-9. Risk Characterization to Corresponding Aquatic and Terrestrial Receptors Assessed for the
Following COUs 129
Table 4-10. Terms and Values Used to Assess Potential Trophic Transfer of TCEP for Terrestrial Risk
Characterization 133
Table 4-11. Environmental RQs by COU with Production Volumes of 2,500 lb/year for Aquatic
Organisms with TCEP Surface Water Concentration (ppb) Modeled by VVWM-PSC 137
Table 4-12. Environmental RQs by COU with Production Volumes of 2,500 lb/year for Aquatic
Organisms with TCEP Pore Water Concentration (ppb) Modeled by VVWM-PSC 138
Table 4-13. RQs Calculated Using Monitored Environmental Concentrations from WQX/WQP 139
Table 4-14. RQs Calculated Using TCEP in Surface Water from Monitored Environmental
Concentrations from Published Literature 139
Table 4-15. Calculated RQs Based on TCEP Soil Concentrations (mg/kg) as Calculated Using Modeled
Data 140
Table 4-16. RQs Calculated Using TCEP Soil Concentrations from Published Literature 140
Table 4-17. RQs for Screening-Level Trophic Transfer of Soil TCEP in Terrestrial Ecosystems Using
EPA's Wildlife Risk Model for Eco-SSLs 141
Table 4-18. RQs for Screening-Level Trophic Transfer of Biosolid TCEP in Terrestrial Ecosystems
Using EPA's Wildlife Risk Model for Eco-SSLs 142
Table 4-19. RQs Calculated with Highest Mean TCEP Soil Concentration (5.89E-03 mg/kg) from
Monitored Values in Published Literature for Screening-Level Trophic Transfer of TCEP
in Terrestrial Ecosystems Using EPA's Wildlife Risk Model for Eco-SSLs 143
Table 4-20. Selected RQs (Highest Fish TCEP Concentrations) Based on Potential Trophic Transfer of
TCEP from Fish to American Mink (Mustela vison) as a Model Aquatic Predator Using
EPA's Wildlife Risk Model for Eco-SSLs 143
Table 4-21. Exposure Scenarios (Production Volume of 2,500 lb TCEP/year) and Corresponding
Environmental Risk for Aquatic Receptors with TCEP in Surface Water and Pore Water
148
Page 11 of 638
-------
Table 4-22. Exposure Scenarios (Production Volume of 2,500 lb TCEP/year) and Corresponding
Environmental Risk for Terrestrial Receptors with TCEP in Soil (Invertebrates) and
Trophic Transfer 149
Table 4-23. TCEP Evidence Table Summarizing Overall Confidence Derived for Trophic Transfer... 158
Table 4-24. TCEP Evidence Table Summarizing Overall Confidence for Environmental Risk
Characterization 160
Table 5-1. Summary for Each OES 166
Table 5-2. Summary of Total Number of Workers and ONUs Potentially Exposed to TCEP for Each
OES 168
Table 5-3. Summary of Inhalation Exposure Results for Workers Based on Monitoring Data for Each
OES 170
Table 5-4. Summary of Inhalation Exposure Results for Workers Based on Exposure Modeling for Each
OES 170
Table 5-5. Summary of Inhalation Exposure Results for ONUs Based on Monitoring Data and Exposure
Modeling for Each OES 172
Table 5-6. Summary of Dermal Retained Dose for Workers Based on Exposure Modeling for Each OES
174
Table 5-7. Summary of Consumer COUs, Exposure Scenarios, and Exposure Routes 179
Table 5-8. CEM 3.2 Model Codes and Descriptions 183
Table 5-9. Crosswalk of COU Subcategories, CEM 3.2 Scenarios, and Relevant CEM 3.2 Models Used
for Consumer Modeling 183
Table 5-10. Summary of Key Parameters for Article Modeling in CEM 3.2 186
Table 5-11. Steady State Air Concentrations and Respirable Particle of TCEP from Consumer Modeling
(CEM 3.2) 187
Table 5-12. Chronic Dermal Average Daily Doses (mg/kg-day) of TCEP from Consumer Article
Modeling for Adults and Children 3 to 6 Years of Age (CEM 3.2) 191
Table 5-13. Chronic Ingestion Average Daily Doses (mg/kg-day) of TCEP from Consumer Article
Modeling for Adults and Infants 1 to 2 Years of Age (CEM 3.2) 192
Table 5-14. Summary of Commercial Paints and Coatings Concentrations and Density of TCEP 194
Table 5-15. Summary of Acute Daily Rate of Consumer Articles Modeled with CEM 3.2 197
Table 5-16. Summary of Chronic Average Daily Doses of Consumer Articles Modeled with CEM 3.2
198
Table 5-17. Summary of Lifetime Average Daily Doses of Consumer Articles Modeled with CEM 3.2
199
Table 5-18. Weight of Scientific Evidence Confidence for Chronic Consumer Exposure Modeling
Scenarios 200
Table 5-19. Sensitivity Analysis for Chronic Consumer Exposure Modeling Scenarios 202
Table 5-20. Summary of Sampling Date for TCEP Weight Fraction Data 205
Table 5-21. Summary of Indoor Monitoring Data of TCEP from U.S. Studies 206
Table 5-22. Summary of Environmental Monitoring Data of TCEP from the Literature for U.S. Studies
209
Table 5-23. Excerpt of Ambient Air Modeled Concentrations for the 2,500 lb Production Volume, High-
End Release Estimate for all COUs at 100 m, Suburban Forest Land Category Scenario
215
Table 5-24. Modeled Incidental Dermal (Swimming) Doses for all COUs for Adults, Youths, and
Children, for the 2,500 lb High-End Release Estimate 216
Table 5-25. Modeled Soil Dermal Doses for the Commercial Use of Paints and Coatings COU, for
Children 217
Page 12 of 638
-------
Table 5-26. 50th Quantile Distances and 30Q5 and Harmonic Mean 50th Quantile Dilution Factors for
Relevant TCEP SIC 220
Table 5-27. Modeled Drinking Water Ingestion Estimates for Diluted Surface Water Concentrations for
Adults for All Industrial and Commercial COUs for the 2,500 lb High-End Release
Estimate 220
Table 5-28. Modeled Drinking Water Ingestion Estimates for Surface Water Concentrations for Adults
for All Industrial and Commercial COUs for the 2,500 lb High-End Release Estimate 221
Table 5-29. Landfill Releases of TCEP from Two Commercial and Industrial OESs 221
Table 5-30. Estimated Average Daily Doses and Lifetime Average Daily Concentrations for Adults from
Groundwater Concentrations by DRAS 222
Table 5-31. Fish Tissue Concentrations Calculated from Modeled Surface Water Concentrations and
Monitoring Data 224
Table 5-32. Adult General Population Fish Ingestion Doses by Scenario Based on a Production Volume
of 2,500 lb/year, High-End Release Distribution, and Modeled Surface Water
Concentrations Based on 50th Percentile Flow of Harmonic Mean 225
Table 5-33. Adult Subsistence Fisher Doses by Scenario Based on a Production Volume of 2,500
lb/year, High-End Release Distribution, and Modeled Surface Water Concentrations
Based on 50th Percentile Flow of Harmonic Mean 226
Table 5-34. Adult Tribal Fish Ingestion Doses by Scenario Based on a Production Volume of 2,500
lb/year, High-End Release Distribution, Modeled Surface Water Concentrations Based on
50th Percentile Flow, and Two Fish Ingestion Rates 227
Table 5-35. Modeled Soil Oral Doses for Soil Concentrations Estimated from Air Deposition and
Biosolids Application for the 2,500 lb High-End Release Estimates 229
Table 5-36. Modeled Incidental Oral (Swimming) Doses for All COUs, for Adults, Youth and Children,
for the 2,500 lb High-End Release Estimate 230
Table 5-37. Concentrations of Foods Found in the Monitoring Literature in ng/g 235
Table 5-38. Human TCEP/BCEP U.S. Biomonitoring Datasets by Population, Type, and Number 238
Table 5-39. Reconstructed Daily Intakes from Creatinine Adjusted Urinary BCEP Concentrations from
NHANES (2013-2014) 239
Table 5-40. General Population Acute Oral Ingestion Estimates for Drinking Water Summary Table 240
Table 5-41. Summary of General Population Chronic Oral Exposures 241
Table 5-42. Summary Acute and Chronic General Population Dermal Exposures 242
Table 5-43. Summary of General Population Inhalation Exposures 242
Table 5-44. Overall Confidence for General Population Exposure Scenarios 243
Table 5-45. Qualitative Assessment of the Uncertainty and Variability Associated with General
Population Assessment 246
Table 5-46. Associations between BCEP (TCEP Metabolite) in Urine of Pregnant Women and Growth
and Gestational Age 279
Table 5-47. Comparison among Studies with Sensitive Neurotoxicity Endpoints Considered for Acute
Exposure Scenarios 290
Table 5-48. Dose-Response Analysis of Selected Studies Considered for Acute Exposure Scenarios . 291
Table 5-49. Comparison among Studies with Sensitive Endpoints Considered for Intermediate Exposure
Scenarios 295
Table 5-50. Dose-Response Analysis of Selected Studies Considered for Intermediate Exposure
Scenarios 296
Table 5-51. Comparison among Studies with Sensitive Endpoints Considered for Chronic Exposure
Scenarios 298
Table 5-52. Dose-Response Analysis of Selected Studies Considered for Chronic Exposure Scenarios
299
Page 13 of 638
-------
Table 5-53. Dose-Response Analysis of Kidney Tumors for Lifetime Exposure Scenarios 301
Table 5-54. Confidence Summary for Human Health Hazard Assessment 306
Table 5-55. Non-cancer HECs and HEDs Used to Estimate Risks 308
Table 5-56. Cancer IUR and CSF Used to Estimate Risks 308
Table 5-57. General Population Acute Drinking Water (Oral Ingestion) Non-cancer Risk Summary .. 309
Table 5-58. Acute Fish Ingestion Non-cancer Risk Summary Based on 50th Percentile Flow of
Harmonic Mean 310
Table 5-59. General Population Chronic Water and Soil Ingestion Non-cancer Risk Summary 311
Table 5-60. Chronic Fish Ingestion Non-cancer Risk Summary 312
Table 5-61. General Population Lifetime Cancer Oral Ingestion Risk Summary Table 313
Table 5-62. Lifetime Cancer Risk Summary for Fish Consumption 314
Table 5-63. General Population Dermal Acute and Chronic Non-cancer Risk Summary 315
Table 5-64. Inhalation Chronic Risk Summary for General Population 316
Table 5-65. General Population Lifetime Cancer Inhalation Risk Summary Table 317
Table 5-66. Exposure Scenarios, Populations of Interest, and Hazard Values 320
Table 5-67. Occupational Risk Summary for 2,500 lb Production Volume 325
Table 5-68. Acute and Chronic Non-cancer Consumer Risk Summary 332
Table 5-69. Lifetime Cancer Consumer Risk Summary 334
Table 5-70. General Population Acute Drinking Water (Oral Ingestion) Non-cancer Risk Summary .. 341
Table 5-71. Acute Fish Ingestion Non-cancer Risk Summary Based on 50th Percentile Flow of
Harmonic Mean 342
Table 5-72. General Population Chronic Water and Soil Ingestion Non-cancer Risk Summary 343
Table 5-73. Chronic Fish Ingestion Non-cancer Risk Summary 344
Table 5-74. General Population Lifetime Cancer Oral Ingestion Risk Summary Table 345
Table 5-75. Lifetime Cancer Risk Summary for Fish Consumption 346
Table 5-76. General Population Dermal Acute and Chronic Non-cancer Risk Summary 347
Table 5-77. Inhalation Chronic Risk Summary for General Population 348
Table 5-78. General Population Lifetime Cancer Inhalation Risk Summary Table 349
Table 5-79. Summary of PESS Considerations Incorporated into the Risk Evaluation 357
Table 5-80. Summary of Detection Frequencies and Sampling Dates for Relevant Consumer Products
Containing TCEP 363
Table 5-81. Suggested Consumer Population Sizes Based on Characterization of Consumer Article
Detection Frequencies 364
Table 5-82. Overall Confidence for Acute, Intermediate, and Chronic Human Health Non-cancer Risk
Characterization for COUs Resulting in Risks 379
Table 5-83. TCEP Evidence Table Summarizing Overall Confidence for Lifetime Human Health Cancer
Risk Characterization for COUs Resulting in Risks 381
Table 6-1. Supporting Basis for the Unreasonable Risk Determination for Human Health (Occupational
COUs) 398
Table 6-2. Supporting Basis for the Unreasonable Risk Determination for Human Health (Consumer
COUs) 400
Table 6-3. Supporting Basis for the Unreasonable Risk Determination for the Environment 401
LIST OF FIGURES
Figure 1-1. TSCA Existing Chemicals Risk Evaluation Process 25
Figure 1-2. TCEP Life Cycle Diagram 26
Figure 1-3. Reported Aggregate TCEP Production Volume (lb) 2012-2020 28
Figure 1-4. TCEP Conceptual Model for Industrial and Commercial Activities and Uses: Potential
Exposure and Hazards 32
Page 14 of 638
-------
Figure 1-5. TCEP Conceptual Model for Consumer Activities and Uses: Potential Exposures and
Hazards 33
Figure 1-6. TCEP Conceptual Model for Environmental Releases and Wastes: General Population
Hazards 34
Figure 1-7. TCEP Conceptual Model for Environmental Releases and Wastes: Ecological Exposures and
Hazards 35
Figure 1-8. Populations Assessed in this Risk Evaluation 36
Figure 1-9. Diagram of the Systematic Review Process 38
Figure 2-1. Transport, Partitioning, and Degradation of TCEP in the Environment 46
Figure 3-1. An Overview of How EPA Estimated Daily Releases for Each OES 51
Figure 3-2. Concentrations of TCEP (ng/m3) in Ambient Air from 2000 to 2019 64
Figure 3-3. Map of Nationwide Measured TCEP Water Concentrations Retrieved from the WQP from
1995 to 2022 68
Figure 3-4. Map Indicating Norman Landfill in Proximity to Tribal Lands 69
Figure 3-5. Groundwater Concentration of TCEP Reported near Twenty-Nine Palms Reservation near
Coachella, California 70
Figure 3-6. Concentrations of TCEP (ng/L) in Surface Water from 1980 to 2017 71
Figure 3-7. Concentrations of TCEP (ng/L) in Precipitation from 1994 to 2014 72
Figure 3-8. Frequency of Nationwide Measured TCEP Surface Water Concentrations Retrieved from the
WQP from 2003 to 2022 73
Figure 3-9. Time Series of Nationwide Measured TCEP Surface Water Concentrations Retrieved from
the WQP from 2003 to 2022 74
Figure 3-10. Concentrations of TCEP (ng/L) in Wastewater from 2001 to 2018 79
Figure 3-11. Concentrations of TCEP (ng/g) in Sediment from 1980 to 2017 80
Figure 3-12. Concentrations of TCEP (ng/L) in the Not Specified Fraction of Groundwater from 1978 to
2017 86
Figure 3-13. Frequency of Nationwide Measured TCEP Groundwater Concentrations Retrieved from the
WQP from 1995 to 2021 87
Figure 3-14. Time Series of Nationwide Measured TCEP Groundwater Concentrations Retrieved from
the WQP from 1995 to 2021 88
Figure 3-15. Concentrations of TCEP (ng/m3) in Indoor Air from 2000 to 2016 93
Figure 3-16. Concentrations of TCEP (ng/m3) in Personal Inhalation in General Population
(Background) Locations from 2013 to 2016 93
Figure 3-17. Concentrations of TCEP (ng/g) in Indoor Dust from 2000 to 2019 96
Figure 4-1. Measured Concentrations of TCEP (ng/g) in Aquatic Species - Fish from 2003 to 2016 .. 101
Figure 4-2. Measured Concentrations of TCEP (ng/g) in Terrestrial Species - Bird from 2000 to 2016
103
Figure 4-3. Measured Concentrations of TCEP (ng/g) in the Wet Fraction of Terrestrial Species - Plant
in Remote (Not Near Source) Locations from 1993 to 1994 104
Figure 4-4. Trophic Transfer of TCEP in Aquatic and Terrestrial Ecosystems 106
Figure 5-1. Approaches Used for Each Component of the Occupational Assessment for Each OES ... 164
Figure 5-2. Consumer Pathways and Routes Evaluated in this Assessment 178
Figure 5-3. Photo of TCEP Label on Wooden Television Stand 182
Figure 5-4. Potential Human Exposure Pathways to TCEP for the General Population 208
Figure 5-5. Direct and Indirect Exposure Assessment Approaches Used to Estimate General Population
Exposure to TCEP 211
Figure 5-6. Modeled Exposure Points for Finite Distance Rings for Ambient Air Modeling (AERMOD)
212
Figure 5-7. General Population Inhalation Concentrations (ppm) by Distance (m) in Log Scale 214
Page 15 of 638
-------
Figure 5-8. Concentrations of TCEP (ng/L) in Drinking Water from 1982 to 2014 218
Figure 5-9. Concentrations of TCEP (ng/g) in the Wet Fraction of Dietary from 1982 to 2018 234
Figure 5-10. Concentrations of TCEP (ng/L) in the Unadjusted Urine from 2015 to 2019 236
Figure 5-11. Concentrations of BCEP (ng/L) in the Creatinine-Adjusted Urine from 2014 to 2019 236
Figure 5-12. Concentrations of BCEP from NHANES data for the U.S. Population from 2011 to 2014
237
Figure 5-13. Concentrations of TCEP (ng/wipe) in Surface Wipes from 2014 to 2018 237
Figure 5-14. Concentrations of TCEP (ng/wipe) in Silicone Wristbands from 2012 to 2015 237
Figure 5-15. Aggregate CADDs for Each Consumer COU, Lifestage 250
Figure 5-16. EPA Approach to Hazard Identification, Data Integration, and Dose-Response Analysis for
TCEP 253
Figure 5-17. Exposure Response Array for Intermediate and Chronic Exposure Durations by Likely
Hazard Outcomes (and Developmental Toxicity) 292
Figure 5-18. Aggregate CADDs for Consumer Use of Textiles in Outdoor Play Structures at Adult,
Youth2, and Youth 1 Lifestages 366
Figure 5-19. Aggregate ADRs for Carpet Back Coating, Childl, and Infant2 Lifestages 367
Figure 5-20. Consumer Modeling Time Series Results for Acoustic Ceilings 372
Figure 5-21. Consumer Modeling Time Series Results for Wood Flooring 373
Figure 5-22. Consumer Modeling Time Series Results for Insulation 373
LIST OF APPENDIX TABLES
Table_Apx B-l. Federal Laws and Regulations 443
Table_Apx B-2. State Laws and Regulations 444
Table_Apx B-3. International Laws and Regulations 445
Table_Apx B-4. Assessment History of TCEP 446
TableApx E-l. PESS Evidence Crosswalk for Increased Exposure 460
TableApx E-2. PESS Evidence Crosswalk for Biological Susceptibility Considerations 464
Table Apx G-l. Invertebrate and Vertebrate Web-ICE Predicted Species that Met Model Selection
Criteria 493
Table Apx G-2. Algal Web-ICE Predicted Species that Met Model Selection Criteria 495
Table Apx G-3. Considerations that Inform Evaluations of the Strength of the Evidence within an
Evidence Stream (i.e., Apical Endpoints, Mechanistic, or Field Studies) 509
Table Apx H-l. Calculated RQs Based on TCEP Surface Water Concentrations (ppb) as Calculated
Using Modeled Data for Annual Air Deposition to Surface Water 511
Table Apx H-2. Environmental RQs by Exposure Scenario with Production Volumes of 2,500 lb/year
with Central Tendency Release Estimates for Aquatic Organisms with TCEP Surface
Water Concentration (ppb) Modeled by VVWM-PSC with 50% Percentile Flow of the
7Q10 512
Table Apx H-3. Environmental RQs by Exposure Scenario with Production Volumes of 2,500 lb/year
with Central Tendency Release Estimates for Aquatic Organisms with TCEP Pore Water
Concentration (ppb) Modeled by VVWM-PSC with 50% Percentile Flow of the 7Q10513
Table Apx H-4. Environmental RQs by Exposure Scenario with Production Volumes of 2,500 lb/year
with Central Tendency Release Estimates for Aquatic Organisms with TCEP Surface
Water Concentration (ppb) Modeled by VVWM-PSC with 90% Percentile Flow of the
7Q10 514
Table Apx H-5. Environmental RQs by Exposure Scenario with Production Volumes of 2,500 lb/year
with Central Tendency Release Estimates for Aquatic Organisms with TCEP Pore Water
Concentration (ppb) Modeled by VVWM-PSC with 90% Percentile Flow of the 7Q10515
Page 16 of 638
-------
TableApx H-6. Environmental RQs by Exposure Scenario with Production Volumes of 2,500 lb/year
with High-End Release Estimates for Aquatic Organisms with TCEP Surface Water
Concentration (ppb) Modeled by VVWM-PSC with 90% Percentile Flow of the 7Q10516
Table Apx H-7. Environmental RQs by Exposure Scenario with Production Volumes of 2,500 lb/year
with High-End Release Estimates for Aquatic Organisms with TCEP Pore Water
Concentration (ppb) Modeled by VVWM-PSC with 90% Percentile Flow of the 7Q10517
Table Apx H-8. Environmental RQs by Exposure Scenario with Production Volumes of 25,000 lb/year
with Central Tendency Release Estimates for Aquatic Organisms with TCEP Surface
Water Concentration (ppb) Modeled by VVWM-PSC with 50% Percentile Flow of the
7Q10 518
Table Apx H-9. Environmental RQs by Exposure Scenario with Production Volumes of 25,000 lb/year
with central tendency release estimates for Aquatic Organisms with TCEP Benthic Pore
Water Concentration (ppb) Modeled by VVWM-PSC with 50% percentile flow of the
7Q10 519
Table Apx H-10. Environmental RQs by Exposure Scenario with Production Volumes of 25,000 lb/year
with central tendency release estimates for Aquatic Organisms with TCEP Surface Water
Concentration (ppb) Modeled by VVWM-PSC with 90% percentile flow of the 7Q10 520
Table Apx H-l 1. Environmental RQs by Exposure Scenario with Production Volumes of 25,000 lb/year
with Central Tendency Release Estimates for Aquatic Organisms with TCEP Benthic
Pore Water Concentration (ppb) Modeled by VVWM-PSC with 90% Percentile Flow of
the 7Q10 521
Table Apx H-12. Calculated RQs Based on TCEP Soils Concentrations (mg/kg) as Calculated Using
Modeled Data for Air Deposition to Soil 522
Table Apx H-13. RQs Based on Potential Trophic Transfer of TCEP in Terrestrial Ecosystems Using
EPA's Wildlife Risk Model for Eco-SSLs (Equation 4-1) 523
Table Apx H-14. RQs Based on Potential Trophic Transfer of TCEP from Fish to American Mink as a
Model Aquatic Predator Using EPA's Wildlife Risk Model for Eco-SSLs (Equation 4-1)
525
Table_Apx 1-1. Body Weight by Age Group 526
Table_Apx 1-2. Fish Ingestion Rates by Age Group 526
Table Apx 1-3. Crosswalk of COU and OES, Abbreviations, and Relevant SIC Codes 531
Table Apx 1-4. Harmonic Mean, 30Q5, 7Q10, and 1Q10 50th Percentile Flows for Relevant TCEP SIC
Codes 532
Table Apx 1-5. Incidental Dermal (Swimming) Modeling Parameters 533
Table Apx 1-6. Incidental Oral Ingestion (Swimming) Modeling Parameters 533
Table Apx 1-7. Ambient Air Release Inputs Utilized for Ambient Air Modeling: IIOAC and AERMOD
Methodology for TCEP 536
Table_Apx 1-8. Settings for Gaseous Deposition 542
Table_Apx 1-9. Settings for Particle Deposition 542
Table Apx 1-10. Description of Daily or Period Average and Air Concentration Statistics 545
Table Apx 1-11. Adult General Population Fish Ingestion Doses by Scenario Based on a Production
Volume of 2,500 lb/year, High-End Release Distribution, and Modeled Surface Water
Concentrations Based on 90th Percentile Flow of Harmonic Mean 547
Table Apx 1-12. Adult Subsistence Fisher Doses by Scenario Based on a Production Volume of 2,500
lb/year, High-End Release Distribution, and Modeled Surface Water Concentrations
Based on 90th Percentile Flow of Harmonic Mean 548
Table Apx 1-13. Adult Tribal Fish Ingestion Doses by Scenario Based on a Production Volume of 2,500
lb/year, High-End Release Distribution, Modeled Surface Water Concentrations Based on
90th Percentile Flow, and Two Fish Ingestion Rates 548
Page 17 of 638
-------
TableApx 1-14. Acute Fish Ingestion Non-cancer Risk Summary Based on 90th Percentile Flow of
Harmonic Mean 550
Table Apx 1-15. Chronic Fish Ingestion Non-cancer Risk Summary Based on 90th Percentile Flow of
Harmonic Mean 551
Table Apx 1-16. Lifetime Cancer Risk Summary for Fish Consumption Based on 90th Percentile Flow
of Harmonic Mean 552
Table_Apx 1-17. Key Chemical Characteristics of TCEP 553
Table Apx 1-18. Data Input Requirements for the Multi-compartment Model 556
Table_Apx 1-19. Mean and Upper Milk Ingestion Rates by Age 558
Table Apx 1-20. Comparison of the Range of Measured and Modeled TCEP Concentrations in Human
Milk 559
Table Apx 1-21. Average Infant Doses via Human Milk Exposure from Maternal Consumer Use
Scenarios 560
Table Apx 1-22. Average Infant Doses from Maternal Workers Based on Mean Milk Intake Rate 562
Table Apx 1-23. Average Infant Doses from Maternal Workers Based on Upper Milk Intake Rate.... 563
Table Apx 1-24. Average Infant Doses via Human Milk Exposure from Maternal General Population
Oral Exposures Based on Mean Milk Intake Rate 564
Table Apx 1-25. Average Infant Doses via Human Milk Exposure from Maternal General Population
Oral Exposures Based on Upper Milk Intake Rate 565
Table Apx 1-26. Average Infant Doses via Human Milk Exposure from Maternal Tribal Fish Ingestion
Based on Mean Milk Intake Rate 566
Table Apx 1-27. Average Infant Doses via Human Milk Exposure from Maternal Tribal Fish Ingestion
Based on Upper Milk Intake Rate 567
Table Apx 1-28. Infant Risks via Human Milk Exposure from Maternal Consumer Use Scenarios .... 568
Table Apx 1-29. Infant Risks via Human Milk Exposure from Maternal Occupational Use Scenarios
Based on Mean Milk Intake Rate 570
Table Apx 1-30. Infant Risks via Human Milk Exposure from Maternal Occupational Use Scenarios
Based on Upper Milk Intake Rate 571
Table Apx 1-31. Infant Risks via Human Milk Exposure from Maternal General Population Oral
Exposures Based on Mean Milk Intake Rate 572
Table Apx 1-32. Infant Risks via Human Milk Exposure from Maternal General Population Oral
Exposures Based on Upper Milk Intake Rate 573
Table Apx 1-33. Infant Risks via Human Milk Exposure from Tribal Maternal Fish Exposures Based on
Mean Milk Intake Rate 574
Table Apx 1-34. Infant Risks via Human Milk Exposure from Tribal Maternal Fish Exposures Based on
Upper Milk Intake Rate 575
Table_Apx 1-35. Variables and Values Used in Sensitivity Analysis 576
Table Apx 1-36. Average Infant Doses Using a Longer TCEP Half-life 578
Table Apx 1-37. Infant Risk Estimate Using a Longer TCEP Half-life 579
Table_Apx 1-38. Input Variables for Chemical of Concern 580
Table_Apx 1-39. Waste Management Unit (WMU) Properties 581
Table_Apx L-l. Evidence Integration for Neurotoxicity 600
Table_Apx L-2. Evidence Integration for Reproductive Effects 605
Table Apx L-3. Evidence Integration for Developmental Effects 608
Table Apx L-4. Evidence Integration Table for Kidney Effects 612
Table Apx L-5. Evidence Integration Table for Liver Effects 614
Table Apx L-6. Evidence Integration Table for Cancer 616
Table_Apx M-l. Results of In Vivo Micronucleus Test 626
Table Apx M-2. Results of Bacterial Reverse Mutation Test in Salmonella typhimurium 627
Page 18 of 638
-------
Table_Apx M-3. TCEP Genotoxicity Studies 629
Table Apx 0-1. Limit of Detection (LOD) and Limit of Quantification (LOQ) Summary for Identified
Air Sampling Analytical Methods for TCEP 638
LIST OF APPENDIX FIGURES
FigureApx F-l. Box and Whisker Plots of Reported Physical and Chemical Property Data Values... 470
FigureApx F-2. Screen Capture of EPI Suite™ Parameters Used to Calculate Fate and Physical and
Chemical Properties for TCEP 473
Figure Apx F-3. EPI Suite™ Level III Fugacity Modeling Graphical Result for TCEP 474
Figure Apx F-4. Screen Capture of OECD Pov and LRTP Screening Tool Parameters Used to Calculate
TCEP's LRTP 475
Figure Apx G-l. SSD Toolbox Interface Showing HC05s and P-Values for Each Distribution Using
Maximum Likelihood Fitting Method Using TCEP's Acute Aquatic Hazard Data for
Vertebrates and Invertebrates (Etterson, 2020) 497
Figure Apx G-2. AICc for the Six Distribution Options in the SSD Toolbox for TCEP's Acute Aquatic
Hazard Data for Vertebrates and Invertebrates (Etterson, 2020) 498
Figure Apx G-3. Q-Q Plots of TCEP Acute Aquatic Hazard Data for Vertebrates and Invertebrates with
the (A) Normal, (B) Logistic, (C) Burr, and (D) Triangular Distributions (Etterson, 2020)
499
Figure Apx G-4. SSD Distribution for TCEP's Acute Hazard Data for Invertebrates and Vertebrates
(Etterson, 2020) 500
Figure Apx G-5. SSD Toolbox Interface Showing HC05s and P-Values for Each Distribution Using
Maximum Likelihood Fitting Method Using TCEP's Acute Aquatic Hazard Data for
Algae (Etterson, 2020) 501
Figure Apx G-6. AICc for the Six Distribution Options in the SSD Toolbox for TCEP's Acute Aquatic
Hazard Data for Algae (Etterson, 2020) 502
Figure Apx G-7. Q-Q Plots of TCEP Acute Aquatic Hazard Data for Algae with the (A) Weibull, (B)
Logistic, (C) Normal, and (D) Burr Distributions (Etterson, 2020) 503
Figure Apx G-8. SSD Distribution for TCEP's Acute Hazard Data for Algae (Etterson, 2020) 504
Figure_Apx G-9. TRV Flow Chart 506
Figure Apx 1-1. Example Tooltips from Media Maps and Time Series Graphs 528
Figure_Apx 1-2. Overview of EPA's Screening-Level Ambient Air Pathway Methodology 535
Figure Apx 1-3. Modeled Exposure Points Locations for Finite Distance Rings 539
Figure_Apx 1-4. Modeled Exposure Points for Area Distance 540
Figure Apx 1-5. Cuticular Resistance as a Function of Vapor Pressure 543
Figure Apx 1-6. Compartments and Exposure Routes for Verner Model 555
Figure_Apx 1-7. Sensitivity Analysis of Model Inputs Measured as Elasticity 576
Figure Apx J-l. Screenshot of Lifetime Average Daily Doses (LADDs) Bar Chart Displaying Tool Tip
for Acoustic Ceiling, Inhalation Estimate 586
Figure Apx J-2. Screenshot of Lifetime Average Daily Doses (LADDs) Bar Chart Displaying Function
to Compare Data on Hover, for Insulation Estimates 586
Figure Apx J-3. Screenshot of Lifetime Average Daily Doses (LADDs) Bar Chart Displaying Bar Chart
that Deselects Inhalation Estimate and Selects Ingestion and Dermal Estimates 587
Figure Apx J-4. Screenshots of Lifetime Average Daily Doses (LADDs) Bar Chart Displaying a
Cropped Subsection of the Figure 587
Figure Apx N-l. Exposure Response Array for Likely and Suggestive Human Health Hazard Outcomes
633
Page 19 of 638
-------
ACKNOWLEDGEMENTS
The Assessment Team gratefully acknowledges the participation, input, and review comments from U.S.
Environmental Protection Agency (EPA or the Agency) Office of Pollution Prevention and Toxics
(OPPT) and Office of Chemical Safety and Pollution Prevention (OCSPP) senior managers and science
advisors. The Agency is also grateful for assistance from EPA contractors Battelle (Contract No. EP-W-
16-017); GDIT (Contract No. HHSN316201200013W); SPS (Contract No. 68HERC20D0021); ERG
(Contract No. 68HERD20A0002); ICF (Contract No. 68HERH22F0259); Versar (Contract No. EP-W-
17-006); and SRC (Contract No. 68HERH19D0022). Special acknowledgement is given for the
contributions of technical experts from EPA's Office of Research and Development (ORD), including
Sandy Raimondo for her review of the Web-ICE methodology in Appendix G.2.1.1.
As part of an intra-agency review, the TCEP Risk Evaluation was provided to multiple EPA Program
Offices for review. Comments were submitted by Office of the Administrator/Office of Children's
Health Protection, Office of Air and Radiation, Office of General Council, ORD, and Office of Water.
Docket
Supporting information can be found in the public docket, Docket IDs EPA-HQ-QPPT-2018-0476 and
EPA-HO-QPPT-2023-0265.
Disclaimer
Reference herein to any specific commercial products, process, or service by trade name, trademark,
manufacturer, or otherwise does not constitute or imply its endorsement, recommendation, or favoring
by the United States Government.
Authors: James Bressette, Xiah Kragie, and Andrea Pfahles-Hutchens (Assessment Leads), Kesha
Forrest and Kara Koehrn (Management Leads), Yousuf Ahmad, Andrea Amati, Edwin Arauz, Amy
Benson, Sarah Gallagher, Lauren Gates, Christopher Green, Leigh Hazel, Keith Jacobs, Rachel
McAnallen, Claudia Menasche, Catherine Ngo, Chloe O'Haire, and Joseph Rappold.
Contributors: Chris Brinkerhoff, Sandy Raimondo, Cecelia Tan, Rony Arauz Melendez, Sarah Au,
Jone Corrales, Kelley Fay, Rebecca Feldman, Patricia Fontenot, Ross Geredien, Annie Jacob, Yadi
Lopez, Grace Kaupas, Kelsey Miller, Sydney Nguyen, Brianne Raccor, Cory Strope, Leora Vegosen,
Jason Wight, and Eva Wong.
Technical Support: Mark Gibson and Hillary Hollinger.
This risk evaluation was reviewed and cleared for release by OPPT and OCSPP leadership.
Page 20 of 638
-------
EXECUTIVE SUMMARY
EPA has evaluated tris(2-chloroethyl) phosphate, or TCEP, under the Toxic Substances Control Act
(TSCA). In this risk evaluation, EPA has determined that TCEP presents an unreasonable risk of
injury to human health and the environment under the conditions of use.
In December 2019, EPA designated TCEP as a high-priority substance for TSCA evaluation and in
August 2020 released the final scope of the risk evaluation. This risk evaluation assesses human health
risk to workers, consumers, and the general population, as well as risk to the environment.
Although U.S. production of TCEP has decreased by about 99 percent since 2014, it is still used
domestically as an additive flame retardant and plasticizer in polymers used in aerospace equipment and
products and as an additive flame retardant in paint and coating manufacturing. In the past, TCEP was
used in many products made in the United States, including fabrics and textiles, some types of foam, and
construction materials—some of which may remain in use today. TCEP may still be found in goods that
are imported into the United States.
Because TCEP is mixed into but not chemically bonded to materials, it can leach out of products and
into the environment. TCEP that is released into the environment from manufacturing processes or
leaching from products primarily ends up in water, sediment, soil, or dust. TCEP may leach out of
materials dumped in landfills and reach groundwater or surface water—in particular from landfills that
do not have an adequate liner system. It can also be released into the air from manufacturing, industrial
processes, and open burning. If TCEP enters the atmosphere, it can be deposited in lakes and rivers
through rain and snowfall. It can be transported long distances via air and water and has been detected in
the Arctic. TCEP concentrations may be even higher indoors than outdoors because it can leach out of
consumer products such as carpets or wooden furniture and attach to household dust. Although TCEP is
persistent in the environment (i.e., it does not easily degrade) and has been detected in organisms such as
fish exposed to TCEP in surface water, it does not appear to bioaccumulate. This is because TCEP does
not accumulate in people or animals at greater concentrations than exist in the environment.
Following the 2023 Draft Risk Evaluation, and in response to public and peer reviewer comments, EPA
made the following key updates to the risk evaluation for TCEP:
1. Revised four existing conditions of use (COUs) to include "automotive articles and replacement
parts containing TCEP" and added one new COU: "Industrial use - paints and coatings - paints
and coatings," in response to the Auto Alliance comments.
2. Revised qualitative assessments of COUs not quantified in the draft risk evaluation based on
similarity of exposure scenarios to COUs that have been quantified.
3. Updated the peer-reviewed literature search in February 2024 to fill data gaps for landfills
(general population, consumer, and environmental hazard), environmental hazard, epidemiology,
and inhalation (human health/animal toxicity).
4. Increased the range of the (Hazardous Waste) Delisting Risk Assessment Software (DRAS)
analysis, bounding leachate concentration by solubility, and increasing loading rates to account
for past disposal practices.
5. Revised concentration(s) of concern (COCs) for chronic aquatic hazards for both sediment and
surface water compartments.
6. Changed the evidence integration/conclusion for developmental toxicity from likely to cause the
effect to suggestive for the effect.
Page 21 of 638
-------
7. Summarized and evaluated epidemiology studies identified in the updated literature search and
by peer reviewers for neurotoxicity as well as kidney, immune/hematological, thyroid, and
lung/respiratory effects, body weight changes, developmental toxicity, and cancer.
8. Updated the human health risk characterization with revised analysis of the Consumer Exposure
Model (CEM) for consumers to account for model updates and minor changes to parameter
inputs.
9. Revised the analysis for Consumer paints and coatings COU, including consumer sensitivity
analysis for minimum weight fraction values and revised dermal and oral ingestion estimates
from soils to include soil concentrations from the Biosolids Screening Tool (BST) analysis.
Unreasonable Risk to Human Health
EPA concluded that kidney cancer, as well as neurological and reproductive systems and non-cancer
effects in kidneys, are the key human health hazards from exposure to TCEP based on evaluation of
human epidemiological studies and laboratory animal testing (see Sections 5.2.3.1 and 5.2.4). The
Agency evaluated the risks of potential neurological effects from acute exposure, reproductive effects
from intermediate and chronic exposures, and increased kidney cancer as relevant from exposure to
TCEP at work, in the home, by breastfeeding, and by eating fish taken from TCEP-contaminated water.
When determining unreasonable risk of TCEP to human health, EPA also accounted for potentially
exposed and susceptible subpopulations—pregnant women, infants exposed through human milk,
children and adolescents, people who experience high exposures or exposures from multiple routes
(such as dermal and inhalation), people who live in fenceline communities near facilities that release
TCEP, firefighters, and people and Tribes whose diets include large amounts of fish (see Section 5.3.3).
Workers with the greatest potential for exposure—both dermal and inhalation—to TCEP are those who
spray TCEP-containing paints or coatings, or workers who are involved in processing a 2-part resin used
in paints, coatings, and polyurethane resin castings for aerospace applications (see Section 5.3.2.1).
Outside the workplace, adults, infants, and children may be most at risk if they breathe or ingest TCEP
released from fabrics, textiles, foam, and wood products and that either attaches to dust or otherwise gets
into indoor air (see Section 5.3.2.2). Infants and children may be at risk if they "mouth" products
containing foam, textiles, or wood that contain TCEP (see Section 5.3.2.3). People who are subsistence
fishers may be at risk if they eat TCEP-contaminated fish. Tribal people for whom fish is important
dietarily and culturally have greater risk than the general population and subsistence fishers due to
increased fish consumption (see Section 5.3.3).
EPA's assessment shows 10 of the 21 conditions of use to significantly contribute to the
unreasonable risks presented by TCEP due to cancer and non-cancer health effects for (1)
workers who handle or apply liquid formulations containing TCEP (due to dermal and inhalation
exposures) as well as workers who fight structural fires; (2) people who breathe or ingest dust
from TCEP released from consumer products; and (3) people who eat large amounts of fish
contaminated with TCEP. For workers, there are certain activities where acute, intermediate, chronic,
and lifetime exposures to TCEP, especially from contact with skin and inhalation of mists generated
during the use of paints and coatings, contribute to unreasonable risk. Outside the work environment,
TCEP presents unreasonable risk to adults, children, and infants with acute, intermediate/chronic, and
lifetime exposure to TCEP—mainly from breathing or ingesting TCEP-containing dust or eating TCEP-
contaminated fish. TCEP presents unreasonable risk to children and infants with acute and
intermediate/chronic exposure from mouthing consumer products that contain TCEP. EPA also assessed
whether breast-feeding infants of mothers from occupational, consumer, general population, subsistence
fisher, and Tribal exposure scenarios were at higher risk than their mothers and determined that they are
not.
Page 22 of 638
-------
Unreasonable Risk to the Environment
EPA assessed TCEP exposures to the aquatic environment when it leaches or is released into water
through the manufacturing, processing, or use of TCEP or TCEP-containing materials. Aquatic hazard
data were available for six fish species, four invertebrate species, five algal species, and from predictive
models for sediment-dwelling organisms. EPA's assessment shows that chronic exposure to TCEP to
fish and to sediment-dwelling organisms under all six conditions of use that were quantitatively
evaluated for risk to the environment significantly contribute to the unreasonable risk. The
Agency, however, has determined that acute exposure to TCEP does not present unreasonable risk to
aquatic organisms (vertebrate and invertebrate species). Data on soil invertebrates and mammals
indicate that acute and chronic exposure to TCEP does not present unreasonable risks to land-dwelling
animals.
Considerations and Next Steps
The COUs evaluated for TCEP are listed in Table 1-1. The following 10 COUs significantly contribute
to the unreasonable risk:
• Manufacturing (import);
• Processing - Incorporation into formulation, mixture, or reaction product - Paint and coating
manufacturing;
• Processing - Incorporation into formulation, mixture, or reaction product - Polymers used in
aerospace equipment and products;
• Processing - Incorporation into article - Aerospace equipment and products and automotive
articles and replacement parts containing TCEP;
• Industrial use - Paints and coatings;
• Commercial use - Paints and coatings;
• Commercial use - Laboratory chemicals;
• Consumer use - Furnishing, cleaning, treatment/care products - Fabric and textile products;
• Consumer use - Furnishing, cleaning, treatment/care products - Foam seating and bedding
products; and
• Consumer use - Construction, paint, electrical, and metal products - Building/construction
materials - Wood and engineered wood products - Wood resin composites.
The following 11 COUs do not significantly contribute to the unreasonable risk:
• Processing - Recycling;
• Distribution in commerce;
• Industrial use - Other use - Aerospace equipment and products and automotive articles and
replacement parts containing TCEP;
• Commercial use - Other use - Aerospace equipment and products and automotive articles and
replacement parts containing TCEP;
• Commercial use - Furnishing, cleaning, treatment/care products - Fabric and textile products;
• Commercial use - Furnishing, cleaning, treatment/care products - Foam seating and bedding
products;
• Commercial use - Construction, paint, electrical, and metal products - Building/construction
materials - Insulation;
• Commercial use - Construction, paint, electrical, and metal products - Building/construction
materials - Wood and engineered wood products - Wood resin composites;
• Consumer use - Paints and coatings, including those found on automotive articles and
replacement parts;
Page 23 of 638
-------
• Consumer use - Construction, paint, electrical, and metal products - Building/construction
materials - Insulation; and
• Disposal.
It is important to emphasize that the estimates of risk in the TCEP evaluation include assumptions and
modeled predictions around which there are varying levels of uncertainty. That stated, the totality of
information and weight of scientific evidence give EPA confidence that under the known, intended, and
reasonably foreseen COUs that are subject to evaluation and regulation under TSCA, TCEP presents
unreasonable risks to human health and the environment under the conditions of use.
Following issuance of this completed risk evaluation for TCEP, EPA will initiate risk management for
TCEP by applying one or more of the requirements under TSCA section 6(a) to the extent necessary so
that TCEP no longer presents an unreasonable risk. The risk management requirements will likely focus
on the COUs significantly contributing to the unreasonable risk. However, under TSCA section 6(a),
EPA is not limited to regulating the specific COUs found to significantly contribute to unreasonable risk
and may select from among a suite of risk management options related to manufacture, processing,
distribution in commerce, commercial use, and disposal to address the unreasonable risk.
Page 24 of 638
-------
1 INTRODUCTION
EPA has evaluated tris(2-chloroethyl) phosphate (TCEP) under the Toxic Substances Control Act
(TSCA). TCEP is primarily used as an additive flame retardant and plasticizer in polymers used in
aerospace equipment and products and as an additive flame retardant in paint and coating
manufacturing. In the past, TCEP was processed in many products made in the United States, including
fabrics and textiles, some types of foam, and construction materials—some of which may still be in use
today. TCEP may also be imported in articles intended for consumer use. Section 1.1 provides
production volume, life cycle diagram (LCD), conditions of use (COUs), and conceptual models used
for TCEP; Section 1.2 includes an overview of the systematic review process; and Section 1.3 presents
the organization of this risk evaluation. Figure 1-1 describes the major inputs, phases, and
outputs/components of the TSCA risk evaluation process, from scoping to releasing the risk evaluation.
Inputs
Existing Laws, Regulations,
and Assessments
Use Document
Public Comments
Public Comments on
Draft Scope Document
• Analysis Plan
• Testing Results
• Data Evaluation Process
Data Integration
• Public Comments on
Draft RE
• Peer Review Comments
on Draft RE
Phase
Outputs
Figure 1-1. TSCA Existing Chemicals Risk Evaluation Process
1.1 Scope of the Risk Evaluation
EPA evaluated risk to human and environmental populations for TCEP. Specifically for human
populations, the Agency evaluated risk to (1) workers and occupational non-users (ONUs) via inhalation
and oral routes; (2) workers via dermal routes; (3) consumers via inhalation, dermal, and oral routes; and
(4) the general population via oral, dermal, and inhalation routes. In this risk evaluation, the general
population includes various subpopulations such as subsistence fishers, Tribal populations, and people
who live in fenceline communities who live near facilities that emit TCEP. For environmental
populations, EPA evaluated risk to (1) aquatic species via water and sediment, and (2) terrestrial species
via air and soil leading to dietary exposure.
1.1.1 Life Cycle and Production Volume
The LCD shown below in Figure 1-2 depicts the COUs that are within the scope of the risk evaluation
during various life cycle stages, including manufacturing, processing, use (industrial, commercial,
consumer), distribution, and disposal. The LCD has been updated since it was included in the TCEP
final scope document (U.S. EPA 2020b) to correspond with minor updates to the COUs. The
information in the LCD is grouped according to the Chemical Data Reporting (CDR) processing codes
and use categories, including functional use codes for industrial uses and product categories for
industrial, commercial, and consumer uses. The CDR Rule under TSCA requires U.S. manufacturers
Page 25 of 638
-------
(including importers) to provide EPA with information on the chemicals they manufacture or import into
the United States. EPA collects CDR data approximately every 4 years with the latest collections
occurring in 2006, 2012, 2016, and 2020, with reporting thresholds of 25,000 lb.
Descriptions of the industrial, commercial, and consumer use categories identified from the CDR are
included in the LCD (Figure 1-2) (U.S. EPA 2016d). The descriptions provide a brief overview of the
use category; the Supplemental Information on Environmental Release and Occupational Exposure
Assessment (U.S. EPA 2024n) contains more detailed descriptions (e.g., process descriptions, worker
activities, process flow diagrams, equipment illustrations) for each manufacture, processing, use, and
disposal category.
Because TCEP is also known to co-occur in formulation with other flame retardants, such as 2,2-
bis(chloromethyl)-propane-l,3-diyltetrakis(2-chloroethyl) bisphosphate (V6), this risk evaluation
evaluates TCEP when it co-occurs with other flame retardants in commercial and consumer products
(e.g., when it co-occurs with V6). However, it does not evaluate the other flame retardants.
MFG/IMPORT
TRIS(2-CHLOROETHYL) PHOSPHATE (TCEP) (CAS RN 115-96-8)
PROCESSING INDUSTRIAL, COMMERCIAL, CONSUMER USES
Incorporation into
Formulation, Mixture, or
Reaction product
(Flame retardart in paint
and coating
manufacturing; Polymers
used in aerospace
equipment and products)
Incorporation into article
(Aerospace equipment and
products and automotive
articles and replacement
parts containing TCEP)
Recycling
Paints and Coatings12
Otheruse1
Aerospace equipment and
products and automotive
articles and replacement
parts containing TCEP;
laboratory chemicals
Furnishing, Cleaning,
Treatment/Care Products12
Fabric and textile products;
foam setting and bedding
products
Construction, Paint,
Electrical, and Metal
Products12
Building/construction
materials not covered
elsewhere; wood resin
composites and insulation
RELEASES and WASTE
DISPOSAL
Disposal
See Conceptual Model for
Environmental Releases
and Wastes
I I Manufacture (Including
— Import)
I I Processing
~ Uses:
1. In
industrial/Commercial
2. Consumer
Figure 1-2. TCEP Life Cycle Diagram
1 Due to lack of reasonably available data, including current CDR data, EPA cannot differentiate between import
and processing sites.
2 See Table 1-1 for additional details on TSCA conditions of use.
As evident in Figure 1-3, import, production volume, and uses of TCEP in the United States have
curtailed in recent years. Although CDR data show production volumes for TCEP in chemical form in
the tens of thousands of pounds from 2012 to 2015, the most recent updated 2020 CDR data showed that
no company reported the manufacture (including import) of TCEP in the United States from 2016 to
2020. However, the reporting threshold for TCEP in CDR is 25,000 lb and some manufacturing could be
Page 26 of 638
-------
occurring below that threshold (U.S. EPA. 2020a).1 The production volumes for TCEP reported to CDR
for years 2012 to 2015 were all from one company, Aceto Corporation, a chemical manufacturer and
supplier importing TCEP in chemical form. Aceto Corporation indicated to EPA that TCEP was
imported and used as a flame retardant for unsaturated polyester resins and for aircraft furniture (U.S.
EPA. 2020b). Note that prior to 2012, production volume in CDR was reported in ranges. From 1986 to
2002, the production volume reported to CDR (previously known as the Inventory Update Rule, or IUR)
was between 1 and 10 million lb. In 2006, the production volume reported was between 500,000 and 1
million lb and in 2011 the production volume was withheld.
To supplement the CDR data, EPA also considered Datamyne import volume information that shows
593 lb of TCEP imported in 2020. Descartes Datamyne is a commercial searchable trade database that
covers the import-export data and global commerce of more than 50 countries (across 5 continents) and
includes cross-border commerce of the United States with over 230 trading partners (Descartes. 2020).
The trade data are gathered from the U.S. Customs Automated Manifest System. Since 2014, total
imports of TCEP in chemical form range in volume over the time from approximately 96,823 lb (in
2014) to 593 lb (in 2020) (Descartes. 2020). Note that for 2014, the Aceto Corporation data is included
in the total production volume for CDR and Datamyne. For 2020, Sigma-Aldrich Corp reported the 593
lb.2
The 2016 CDR reporting data and Datamyne import volume data for TCEP in chemical form are
provided in Figure 1-3. TCEP imported in articles is not captured in these data. Note, EPA only recently
added TCEP to the Toxics Release Inventory (TRI) with the first year of reporting from facilities due
July 1, 2024. As of September 2024, there are no TRI entries for TCEP.
1 Note that because CDR generally does not include information on impurities or manufacturing solely in small quantities for
research and development, and because small manufacturers are exempt from 2020 CDR reporting, some manufacturing
could be occuring at small manufacturers. However, EPA does not consider domestic manufacturing of TCEP to be
reasonably foreseeable. Lastly, TCEP imported in articles would not be captured in CDR.
2 Due to the nature of Datamyne data, some shipments containing TCEP may be excluded due to being categorized under
other names that were not included in the search terms. There also may be errors in the data that prevent shipment records
containing the chemical from being located. Datamyne does not include articles/products containing the chemical unless the
chemical name is included in the description; however, based on descriptions provided on the bills of lading. Figure 1-3
provides an estimate of the volume of TCEP imported as the chemical (not in an identified product) from 2012 to 2020.
Page 27 of 638
-------
Reported Aggregate TCEP Production Volume (lbs.) by Year
2012 2013 2014 2015 2016 2017 2018 2019 2020
¦ CDR Production Volume ¦ Datamyne Import Volume
Figure 1-3. Reported Aggregate TCEP Production Volume (lb) 2012-2020
Note: CDR data for the 2016 reporting period is available via ChemView. Because of an ongoing CBI
substantiation process required by TSCA, the CDR data available in this risk evaluation is more specific than
currently provided in ChemView (U.S. EPA. 2019a). For 2014, Aceto Corporation's production volume is
included in both the CDR data and the Datamyne data.
Given the uncertainties in the current production volume for TCEP, EPA used two production volumes
in its analyses for this risk evaluation: 2,500 and 25,000 lb. The 2,500 lb production volume is used as a
more realistic estimate reflecting current production volumes, while 25,000 lb is used as an upper bound
estimate based on the 2020 CDR reporting threshold. There are several reasons why EPA considers
2,500 lb to be a more realistic production volume. First, the decreasing aggregate TCEP production
volumes according to CDR and Datamyne, as shown in Figure 1-3, suggest that the production volume
is now somewhere below the 2020 CDR reporting threshold of 25,000 lb, with Datamyne showing 593
lb of TCEP imported in 2020 and generally the most recent Datamyne information (2017-2020) in the
low thousands of pounds or lower. Additionally, EPA received public comments (EPA-HQ-OPPT-2018-
0476-0041) on the final scope document (U.S. EPA 2020b) confirming industry's transition away from
the domestic use of TCEP.
Communication with industry further supported the declining use of TCEP as many companies have
since discontinued or reformulated products that contained TCEP, even though TCEP is still in use for
several commercial and consumer COUs (EPA-HQ-OPPT-2018-0476-0056). However, there is no
federal ban on the manufacture, process, or use of TCEP that would prevent production volumes from
increasing again (see Appendix B for the regulatory history of TCEP). Therefore, EPA used these two
160,000
140,000
120,000
100,000
CO
"O
I 80,000
o
CL
60,000
40,000
20,000
Page 28 of 638
-------
production volumes to characterize what is possible and what is realistic given reasonably available
information. Given EPA's research, the 25,000 lb upper bound production volume is believed to be an
overestimate of current production volumes in the United States. For these reasons, the 2,500 lb
production volume is used throughout this risk evaluation as EPA has more confidence that it is
reflective of current production volumes. Estimates using the upper bound of 25,000 lb are presented in
appendices and supplemental files.
1.1.2 CPUs Included in the Risk Evaluation
The Final Scope of the Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) CASRN115-96-8
(U.S. EPA. 2020b) identified and described the life cycle stages, categories and subcategories that
comprise COUs that EPA planned to consider in the risk evaluation. All COUs for TCEP included in
this risk evaluation are reflected in the LCD (Figure 1-2) and conceptual models (Section 1.1.2.1). Table
1-1 below presents all COUs for TCEP.
In this risk evaluation, EPA made edits to the COUs listed in the final scope document. These edits
reflect EPA's improved understanding of the COUs based on further outreach and public comments
received, which have been added to the reference(s) column of Table 1-1. Changes include removing
"flame retardant" as the exclusive functional use in the processing conditions of use; editing industrial
and commercial use in "aircraft interiors and products" to "aerospace equipment and products"; and
improved the description of the COU to avoid using the "products not covered elsewhere" description
from the CDR reporting codes. EPA did not receive public comments on additional commercial uses
that fall into the "Other use" category aside from laboratory chemicals, the Agency removed "e.g.,"
from the COU, "Commercial use - other use - e.g., laboratory chemicals."
All COUs assessed in this Risk Evaluation are considered on-going uses. However, there are several
COUs for which part of the life cycle has ceased, such as manufacturing (including import) and
processing. However, other parts of the life cycle including recycling, commercial or consumer use, and
disposal are on-going. These COUs are identified in Table 1-1 and include four COUs for commercial
use and five COUs for consumer use.
EPA reviewed comments submitted for docket EPA-HQ-OPPT-2023-2012-0001 Significant New Use
Rule for Certain Non-ongoing Uses; Flame Retardants, for certain non-ongoing uses of TCEP and made
corresponding edits to Table 1-1, including the addition of a new COU, Industrial use - Paints and
coatings, and "automotive articles and replacement parts containing TCEP" to some processing,
industrial, and commercial uses. Appendix D contains a description of the COUs.
Page 29 of 638
-------
Table 1-1. Cont
itions of Uses in the Ris
i Evaluation for TCE
>
Life Cycle
Stage"
Category6
Subcategoryc
Reference(s)
Manufacturing
Import
Import
U.S. EPA (2016d)
Processing
Incorporation into
formulation, mixture, or
reaction product
Paint and coating
manufacturing
U.S. EPA (2019a); Duratec (2018);
PPG (2010); PPG (2016); U.S. EPA
(2017c)
Flame Control Coatings_meeting
memo
Incorporation into
formulation, mixture, or
reaction product
Polymers used in
aerospace equipment
and products
EPA-HQ-OPPT-2018-0476-0015;
EPA-HQ-OPPT-2018-0476-0012; BJB
Enterprises (2017); EPA-HO-OPPT-
2018-0476-0045; Summary of email
exchanges
Incorporation into article
Aerospace equipment
and products and
automotive articles
and replacement parts
containing TCEP
EPA-HQ-OPPT-2018-0476-0006;
EPA-HQ-OPPT-2018-0476-0045;
Boeing meeting memo; EPA-HQ-
OPPT-2023-0265-0043; EPA-HQ-
OPPT-2023-0012-0010
Recycling
Recycling
U.S. EPA (2019a)
Distribution in
Commerce
Distribution in
commerce
Distribution in
commerce
Industrial Use
Other use
Aerospace equipment
and products and
automotive articles
and replacement parts
containing TCEP
EPA-HQ-OPPT-2018-0476-0006;
Boeing meeting memo; EPA-HQ-
OPPT-2023-0265-0043; EPA-HQ-
OPPT-2023-0012-0010
Paints and coatings
Paints and coatings
EPA-HQ-OPPT-2023 -0265 -0043;
EPA-HQ-OPPT-2023-0012-0010
Commercial
Use
Other use
Aerospace equipment
and products and
automotive articles
and replacement parts
containing TCEP
EPA-HQ-OPPT-2018-0476-0006;
EPA-HQ-OPPT-2023 -0265 -0043;
EPA-HQ-OPPT-2023-0012-0010
Paints and coatings
Paints and coatings
U.S. EPA (2019a); Alliance for
Automotive Innovation
Laboratory chemicals
Laboratory chemical
TCI America (2018)
Furnishing, cleaning,
treatment/care products
Fabric and textile
products''
EPA-HQ-OPPT-2018-0476-0015
Furnishing, cleaning,
treatment/care products
Foam seating and
bedding products1'
Stapleton et al. (2011); Stapleton &
Hammel meeting memo
Construction, paint,
electrical, and metal
products
Building/construction
materials - insulation1'
EPA-HQ-OPPT-2018-0476-0015;
EPA-HQ-OPPT-2018-0476-0041; EC
(2009). citina IARC (1990)
Page 30 of 638
-------
Life Cycle
Stage"
Category6
Subcategoryc
Reference(s)
Commercial
Use
Construction, paint,
electrical, and metal
products
Building/construction
materials - wood and
engineered wood
products - wood resin
composites''
EC (2009). citina IARC (1990).
OECD (2006) and IPCS (1998)
Paints and coatings
Paints and coatings1',
including those found
on automotive articles
and replacement parts
U.S. EPA (2019a); Alliance for
Automotive Innovation; EPA-HQ-
OPPT-2023-0265-0043; EPA-HQ-
OPPT-2023-0012-0010
Furnishing, cleaning,
treatment/care products
Fabric and textile
products1'
EPA-HQ-OPPT-2018-0476-0015
Consumer Use
Furnishing, cleaning,
treatment/care products
Foam seating and
bedding products1'
Stapleton et al. (2011); Stapleton &
Hammel meeting memo
Construction, paint,
electrical, and metal
products
Building/construction
materials - insulation1'
EPA-HQ-OPPT-2018-0476-0015;
EPA-HQ-OPPT-2018-0476-0041; EC
(2009). citina IARC (1990)
Construction, paint,
electrical, and metal
products
Building/construction
materials -wood and
engineered wood
products - wood resin
composites1'
EC (2009). citina IARC (1990).
OECD (2006). and IPCS (1998)
Disposal
Disposal
Disposal6
11 Life Cycle Stage Use Definitions (40 CFR 711.3)
- "Industrial Use" means use at a site at which one or more chemicals or mixtures are manufactured (including
imported) or processed.
- "Commercial Use" means the use of a chemical or a mixture containing a chemical (including as part of an
article) in a commercial enterprise providing saleable goods or services.
- "Consumer Use" means the use of a chemical or a mixture containing a chemical (including as part of an
article, such as furniture or clothing) when sold to or made available to consumers for their use.
- Although EPA has identified both industrial and commercial uses here for purposes of distinguishing scenarios
in this document, the Agency interprets the authority over "any manner or method of commercial use" under
TSCA section 6(a)(5) to reach both.
h These categories of COU appear in the LCD, reflect CDR codes, and broadly represent COUs of TCEP in
industrial and/or commercial settings and for consumer uses.
c These subcategories reflect more specific COUs of TCEP.
d Domestic manufacturing and processing for these COUs has ceased.
'' This COU includes associated disposal of those COUs for which domestic manufacturing and/or processing have
ceased.
1.1.2.1 Conceptual Models
The conceptual model in Figure 1-4 presents the exposure pathways, exposure routes, and hazards to
human populations from industrial and commercial activities and uses of TCEP. Figure 1-5 presents the
conceptual model for consumer activities and uses, Figure 1-6 presents general population exposure
pathways and hazards for environmental releases and wastes, and Figure 1-7. presents the conceptual
model for ecological exposures and hazards from environmental releases and wastes.
Page 31 of 638
-------
INDUSTRIAL AND COMMERCIAL
ACTIVITIES/USES
Manufacture (incl.
Import)
Processing:
Incorporation into
formulation, mixture,
or reaction product
Incorporation into
article
Recycling
Painting and Coatings
Other use
Furnishing, Cleaning,
Treatment/Care
Products
Construction, Paint,
Electrical,and Metal
Products
Waste Handling, Treatment,
and Disposal
1
EXPOSURE PATHWAY
EXPOSURE ROUTE
POPULATIONS
HAZARDS
-------
CONSUMER ACTIVITIES/USES EXPOSURE PATHWAYS EXPOSURE ROUTES POPULATIONS HAZARDS
Figure 1-5. TCEP Conceptual Model for Consumer Activities and Uses: Potential Exposures and Hazards
The conceptual model presents the exposure pathways, exposure routes, and hazards to human populations from consumer activities and uses of TCEP.
Page 33 of 638
-------
RELEASES AND WASTES FROM
INDUSTRIAL/COMMERCIAL/
CONSUMER USES
Industrial Hre-
Treatment or
Industrial
! wyvT
Indirect discharge
~
EXPOSURE PATHWAYS
EXPOSURE
ROUTES
POPULATIONS
Wastewater or
Liquid Wastes
Solid Wastes
Liquid Wastes
Emissions to
Air
POTW
^ Underground
Injection
Hazardous and
-> Municipal
Waste Landfill
Hazardous and
Municipal
Waste
Incinerators
Off-site Waste
Transfer
Recycling
Other
Treatment
HAZARDS
Hazards Potentially
Associated with
Acute and/or Chronic
Exposures
Figure 1-6. TCEP Conceptual Model for Environmental Releases and Wastes: General Population Hazards
The conceptual model presents the exposure pathways, exposure routes, and hazards to human populations from releases and wastes from industrial,
commercial, and/or consumer uses of TCEP.
Page 34 of 638
-------
RELEASES AND WASTES FROM
INDUSTRIAL / COMMERCIAL /
CONSUMER USES
EXPOSURE PATHWAYS
POPULATIONS
HAZARDS
Figure 1-7. TCEP Conceptual Model for Environmental Releases and Wastes: Ecological Exposures and Hazards
The conceptual model presents the exposure pathways, exposure routes, and hazards to environmental populations from releases and wastes from
industrial, commercial, and/or consumer uses of TCEP.
Page 35 of 638
-------
1.1.3 Populations Assessed
Based on the conceptual models presented in Section 1.1.2.1, Figure 1-8 presents the human and
ecological populations assessed in this risk evaluation. Specifically for humans, EPA evaluated risk to
workers and ONUs via inhalation routes and risk to workers via dermal routes; risk to consumers via
inhalation, dermal, and oral routes; risk to the general population via oral, dermal, and inhalation routes.
For environmental populations, EPA evaluated risk to aquatic species via water and sediment, and risk
to terrestrial species via air, soil, and water leading to dietary exposure. Human health risks were
evaluated for acute, intermediate, chronic, and lifetime exposure scenarios as appropriate, and
environmental risks were evaluated for acute and chronic exposure scenarios, as applicable based on
reasonably available exposure and hazard data as well as the relevant populations for each. All
consumers of products containing TCEP were considered users of those products, and bystanders were
not assessed separately because all the consumer COUs assessed were article scenarios. For the purposes
of article exposures, consumers and bystanders are considered the same.
Environmental
Aquatic
Terrestrial
Surface
Water
Sediment
Soil
Air
Surface
water
AquaticSpecies
\
. /
Terrestrial
Species
Human Health
(includes PESS*)
Occupational
(includes adolescents and
women of reproductive
age)
\J
Consumer
(includes children)
Workers
Inhalation
Dermal
ONUs
1 Inhalation
Users
Inhalation
Ingestion
Dermal
*PESS: Potentially exposed orsusceptible subpopulations
r a
General
Population
(includes fenceline)
v y
All lifestages
Inhalation
Ingestion
Dermal
Figure 1-8. Populations Assessed in this Risk Evaluation
1.1.3.1 Potentially Exposed or Susceptible Subpopulations
TSCA section 6(b)(4)(A) requires that risk evaluations "determine whether a chemical substance
presents an unreasonable risk of injury to health or the environment, without consideration of costs or
other non-risk factors, including an unreasonable risk to a potentially exposed or susceptible
subpopulation identified as relevant to the risk evaluation by the Administrator, under the conditions of
use." TSCA section 3(12) states that "the term 'potentially exposed or susceptible subpopulation'' means
a group of individuals within the general population identified by the Administrator who, due to either
greater susceptibility or greater exposure, may be at greater risk than the general population of adverse
health effects from exposure to a chemical substance or mixture, such as infants, children, pregnant
women, workers, or the elderly."
This risk evaluation considers potentially exposed or susceptible subpopulations (PESS) throughout the
human health risk assessment (see Section 5). Considerations related to PESS can influence the selection
Page 36 of 638
-------
of relevant exposure pathways, the sensitivity of derived hazard values, the inclusion of particular
human populations, and the discussion of uncertainties throughout the assessment. Evaluation of the
qualitative and quantitative evidence for PESS begins as part of the systematic review process, where
any available relevant published studies and other data are identified. If adequate and complete, this
evidence informs the derivation of exposure estimates and human health hazard endpoints/values that
are protective of PESS.
EPA has identified a list of specific PESS factors that may contribute to a group having increased
exposure or biological susceptibility, such as lifestage, occupational and certain consumer exposures,
nutrition, and lifestyle activities. For TCEP, the Agency identified how the risk evaluation addressed
these factors as well as any remaining uncertainties. For the TCEP risk evaluation, EPA accounted for
the following PESS groups: infants exposed through human milk from exposed individuals, children and
male adolescents who use consumer articles or among the exposed general population, subsistence
fishers, Tribal populations, pregnant women, workers and consumers who experience aggregated or
sentinel exposures, people living in fenceline communities near facilities that emit TCEP, and
firefighters. See Section 5.3.3 and Appendix D for details related to this analysis.
1.2 Systematic Review
The U.S. EPA's Office of Pollution Prevention and Toxics (EPA/OPPT) applies systematic review
principles and approaches in the development of risk evaluations. TSCA section 26(h) requires EPA to
use scientific information, technical procedures, measures, methods, protocols, methodologies, and
models consistent with the best available science and base decisions under section 6 on the weight of
scientific evidence. Systematic review supports the risk evaluation through data searching, screening,
evaluation, extraction, and evidence integration and is used to develop the exposure and hazard
assessments based on reasonably available information. EPA defines "reasonably available information"
to mean information that EPA possesses or can reasonably generate, obtain, and synthesize for use in
risk evaluations, considering the deadlines for completing the evaluation (40 CFR 702.33).
In response to comments received by the National Academies of Sciences, Engineering, and Medicine
(NASEM), TSCA Scientific Advisory Committee on Chemicals (SACC) and public, EPA developed the
Draft Systematic Review Protocol Supporting TSCA Risk Evaluations for Chemical Substances (U.S.
EPA. 2021a) (also referred to as the "2021 Draft Systematic Review Protocol") to describe systematic
review approaches implemented in TSCA risk evaluations. In response to recommendations for
chemical specific systematic review protocols, the Risk Evaluation for Tris(2-chloroethyl) Phosphate
(TCEP) - Systematic Review Protocol (U.S. EPA. 2024p) (also referred to as the "TCEP Systematic
Review Protocol") describes clarifications and updates to approaches outlined in the 2021 Draft
Systematic Review Protocol that reflect NASEM, SACC, and public comments as well as chemical-
specific risk evaluation needs. For example, EPA has updated the data quality evaluation process and
will not implement quantitative methodologies to determine both metric and overall data or information
source data quality determinations. Screening decision terminology (e.g., "met screening criteria" as
opposed to "include") was also updated for greater consistency and transparency and to more
appropriately describe when information within a given data source met discipline-specific title and
abstract or full-text screening criteria. Additional updates and clarifications relevant for TCEP data
sources are described in greater detail in the TCEP Systematic Review Protocol (U.S. EPA. 2024p).
The systematic review process is briefly described in Figure 1-9 below. Additional details regarding
these steps are available in the 2021 Draft Systematic Review Protocol (U.S. EPA. 2021a). Literature
inventory trees and evidence maps for each discipline (e.g., human health hazard) displaying results of
Page 37 of 638
-------
the literature search and screening, as well as sections summarizing data evaluation, extraction, and
evidence integration are included in the TCEP Systematic Review Protocol (U.S. EPA. 2024p).
• Based on the
approach
described in the
Literature
Search Strategy
documents.
Data Search
• Title/abstract and
full-text screening
based on pre-
defined
inclusion/exclusion
criteria.
~
Data Screen
• Evaluateand
document the
quality of studies
based on pre-
defined criteria.
Data
Evaluation
• Extract relevant
information based
on pre-defined
templates.
Data
Extraction
=1
1
• Evaluate results
both within and
across evidence
streams to develop
weight of the
scientific evidence
conclusions.
Evidence
Integration
[A
Figure 1-9. Diagram of the Systematic Review Process
EPA used reasonably available information, defined in Title 40 Code of Federal Regulations (40 CFR)
Section 702.33, in a fit-for-purpose approach, to develop a risk evaluation that relies on the best
available science and is based on the weight of scientific evidence in accordance with TSCA sections 6
and 26. EPA reviewed reasonably available information and evaluated the quality of the methods and
reporting of results of the individual studies using the evaluation strategies described in the 2021 Draft
Systematic Review Protocol (U.S. EPA. 2021a) and the TCEP Systematic Review Protocol (U.S. EPA.
2024p).
EPA also identified key assessments conducted by other EPA programs and other U.S. and international
organizations. Depending on the source, these assessments may include information on COUs (or the
equivalent), hazards, exposures, and potentially exposed or susceptible subpopulations. Some of the
most pertinent assessments that were consulted for TCEP include the following:
• U.S. EPA's 2009 Provisional Peer-Reviewed Toxicity Values (PPR1V) for Tris(2-
chloroethvDphosphate (TCEP) (CASRN115-96-8):
• 2009 European Union Risk Assessment Report: CAS: 115-96-8: Tris (2-chloroethyl) phosphate,
TCEP:
• Environment Canada and Health Canada's 2009 Screening Assessment for the Challenge
Ethanol, 2-chloro- phosphate (3:1) (Tris(2-chlrorethyl) phosphate fTCEPl):
• Australia's 2016 Ethanol 2-chloro- phosphate (3:1): Human health tier II assessment:
• Australia's 2017 Ethanol 2-chloro- phosphate (3:1): Human health tier III assessment:
• ATSDR's 2012 ToxicologicalProfile for Phosphate Ester Flame Retardants:
• NTP's 1991 Technical Report on Toxicology and Carcinogenesis Studies of Tris(2-chloroethyl)
Phosphate (CASRN 115-96-8) in F344 N Rats and B6C3F1 Mice (Gavage Studies) : and
• IARC's 1999 Monographs on the Evaluation of Carcinogenic Risks to Humans Volume 71.
An update to the peer literature search to capture information published since 2019 was performed in
February 2024. The search criteria, resources, and search strings used in the literature search update for
TCEP can be found in Section 3.1 of the TCEP Systematic Review Protocol (U.S. EPA. 2024p). After
the search update was complete, additional filtering steps were performed to produce subsets of
literature relevant to targeted information areas to fill known data gaps, namely: landfills (general
population, consumer, and environmental hazard; n = 11); environmental hazard (n = 13); epidemiology
(n = 73); and inhalation (human health/animal toxicity; n = 50). Subsequent discipline-specific title,
abstract and full-text screening criteria (e.g., PECO/PESO/RESO) were leveraged for further
Page 38 of 638
-------
prioritization of references for use in the risk evaluation, as described in Section 4 of the TCEP
Systematic Review Protocol (U.S. EPA. 2024pV
1.3 Organization of the Risk Evaluation
This risk evaluation for TCEP includes five additional sections, a list of REFERENCES, and several
APPENDICES. Section 2 summarizes basic physical-chemical characteristics as well as the fate and
transport of TCEP. Section 3 includes an overview of releases and concentrations of TCEP in the
environment. Section 4 provides a discussion and analysis of the environmental risk assessment,
including the environmental exposure, hazard, and risk characterization based on the COUs for TCEP.
Section 5 presents the human health risk assessment, including the exposure, hazard, and risk
characterization based on the COUs. Section 5 also includes a discussion of PESS based on both greater
exposure and/or susceptibility, as well as a description of aggregate and sentinel exposures. Sections 4
and 5 both discuss any assumptions and uncertainties and how they impact the risk evaluation. Finally,
Section 6 presents EPA's risk evaluation of whether the chemical presents an unreasonable risk to
human health or the environment under the assessed COUs.
Appendix A provides a list of abbreviations and acronyms as well a glossary of select terms used
throughout this risk evaluation. Appendix B provides a brief summary of the federal, state, and
international regulatory history of TCEP. Appendix C lists all separate supplemental files associated
with this risk evaluation, which can be accessed through hyperlinks included in the references. All
subsequent appendices include more detailed analysis and discussion than are provided in the main body
of this risk evaluation for TCEP.
Page 39 of 638
-------
2 CHEMISTRY AND FATE AND TRANSPORT
Physical and chemical properties determine the behavior and characteristics of a chemical that inform its
condition of use, environmental fate and transport, potential toxicity, exposure pathways, routes, and
hazards. Environmental fate and transport include environmental partitioning, accumulation,
degradation, and transformation processes. Environmental transport is the movement of the chemical
within and between environmental media, such as air, water, soil, and sediment. Transformation or
degradation occur through reaction of the chemical in the environment. Thus, understanding the
environmental fate of TCEP informs the determination of the specific exposure pathways, and potential
human and environmental populations that EPA considered in this risk evaluation.
2.1 Physical and Chemical Properties
EPA gathered and evaluated physical and chemical property data and information according to the
process described in the 2021 Draft Systematic Review Protocol (U.S. EPA 2021a). During the
evaluation of TCEP, EPA considered both measured and estimated physical and chemical property
data/information summarized in Table 2-1, as applicable. More details are given in Appendix F.l.
Information on the full, extracted dataset is available in the supplemental file Risk Evaluation for Tris
(2-chloroethyl) Phosphate (TCEP) - Systematic Review of Data Quality Evaluation and Data Extraction
Information for Physical and Chemical Properties (U.S. EPA. 2024v).
TCEP is a clear, transparent liquid with a slight odor (DOE. 2016; U.S. EPA. 2015b; ECB. 2009; Lewis
and Hawlev. 2007; Weil. 2001) and low viscosity (IARC. 1990). TCEP is a chlorinated phosphate ester
that is used as a flame-retardant additive and plasticizer. It melts around -55 °C and begins to
decompose at 320 °C (DOE. 2016; U.S. EPA 2015b; Toscano and Coleman. 2012; ECB. 2009; IARC.
1990). TCEP is soluble in water with a water solubility of 7,820 mg/L at 20 °C (ECB. 2009) and
hydrophilic with a logarithmic octanol:water partition coefficient (log Kow) value of 1.78 (ECB. 2009).
TCEP is reported to have low volatility with a vapor pressure of 0.0613 mm Hg at 25 °C (Dobry and
Keller. 1957) and a boiling point of 330 °C (U.S. EPA 2019b; DOE. 2016; U.S. EPA 2015a; Havnes.
2014; Toscano and Coleman. 2012). However, it will become more volatile when the temperature
increases (0.5 mm Hg at 145 °C) (Toscano and Coleman. 2012; NTP. 1992). Because of its high boiling
point, low volatility, and a Henry's Law constant of 2.945x 10~6 atm-m3/mol at 25 °C (U.S. EPA.
2017a). TCEP is categorized as a semi-volatile organic compound (SVOC) (ECHA. 2018; TERA.
2015).
Table 2-1. Physical and Chemical Properties of TCEP
Property
Selected Value"
Reference(s)
Overall Quality
Determination6
Molecular formula
CeHi.CbCUP
NLM (2019)
High
Molecular weight
285.49 g/mol
Havnes(2014)
High
Physical form
Clear, transparent
liquid with slight
odor
Weil (2001); Lewis and Hawlev (2007);
ECB (2009); U.S. EPA (2015b); DOE
(2016)
High
Melting point
-55 °C
Toscano and Coleman (2012). as cited in
ATSDR (2012). NLM (2015). U.S. EPA
(2015a). U.S. EPA (2015b). and DOE
(2016)
High
Page 40 of 638
-------
Property
Selected Value"
Reference(s)
Overall Quality
Determination6
Boiling point
330 °C
Toscano and Coleman (2012); Havnes
(2014); ATSDR (2012); NLM (2015); U.S.
EPA (2015a); DOE (2016); U.S. EPA
(2019b)
High
Density
1.39 g/cm3 at 25 °C
Toscano and Coleman (2012); Havnes
(2014); DOE (2016)
High
Vapor pressure
0.0613 mm Hg at 25
°C
Dobrv and Keller (1957) as cited in
Verbrueeen et al. (2005). ATSDR (2012).
NLM (2015). U.S. EPA (2015a). and U.S.
EPA (2019b)
High
Vapor density
9.8 (air = 1)
ILO (2019)
High
Water solubility
7,820 mg/L at 20 °C
ECB (2009) as cited in Verbruaaen et al.
(2005). EC (2009). U.S. EPA (2015b). U.S.
EPA (2015a). and NLM (2015)
High
Logarithmic
octanol: water partition
coefficient (log Kow)
1.78
ECB (2009) as cited in Verbruaaen et al.
(2005). EC (2009). U.S. EPA (2015b). U.S.
EPA (2015a). and NLM (2015)
High
Logarithmic octanol:air
partition coefficient
(log Koa)
7.86-7.93
Okeme et al. (2020)
High
Henry's Law constant
2.945E-06
atm m3/mol at 25 °C
(calculated)
U.S. EPA (2017a)
High
Flash point
225 °C (closed cup)
)0) as cited in U.S. EPA (2015a)
High
Autoflammability
480 °C
ECB (2009); ILO (2019)
Medium
Viscosity
45 cP at 20 °C
IARC (1990)
High
Refractive index
1.4721 at 20 °C
Havnes (2014) as cited in NLM (2015)
High
11 Measured unless otherwise noted
h "Overall Quality Determination" apply to all references listed in this table
Page 41 of 638
-------
2.2 Environmental Fate and Transport
Environmental Fate and Transport (Section 2.2):
Key Points
EPA evaluated the reasonably available information to characterize the environmental fate and
transport of TCEP, the key points are summarized below:
• TCEP exists in both gaseous and particle phases under environmentally relevant conditions
and partitions to organic carbon in the air. TCEP is not expected to undergo significant direct
photolysis, but TCEP in the gaseous phase will rapidly degrade in the atmosphere (ti/2 = 5.8
hours).
• TCEP is not expected to undergo abiotic degradation processes such as photolysis and
hydrolysis in aquatic environments under environmentally relevant conditions. However,
TCEP's rate of hydrolysis is highly dependent on pH and temperature conditions.
• TCEP does not biodegrade in water under aerobic conditions but will volatilize from surface
water despite its low Henry's Law constant (2.945xlCT6 atmm3/mol at 25 °C).
• TCEP can be transported to sediment from overlying surface water through advection and
dispersion of dissolved TCEP and deposition of suspended solids containing TCEP. Based on
its log Koc values (3.23-3.46), TCEP is expected to partition to organic matter in suspended
and benthic sediments. However, TCEP may partition between surface water and sediments
to varying degrees because of its high water solubility (7,820 mg/L at 20 °C).
• TCEP accumulation in soil is unlikely because of its log Koc values (2.08-2.52). Due to its
high water solubility and its low Henry's Law constant, TCEP in moist soil will both migrate
to groundwater and volatilize.
• TCEP will be minimally removed via conventional drinking water and wastewater treatment
and will be retained in wastewater effluents with a low fraction being adsorbed into sludge.
• TCEP has been detected in surface water and groundwater samples; point sources include
wastewater effluents and landfill leachates.
• TCEP has been detected in surface water, air, and snow in remote locations with no known
source of releases but is known to undergo long-range transport through atmospheric, plastic
debris, and other natural processes.
• TCEP does not bioaccumulate in aquatic fish but may in benthic fish. When TCEP
concentrations are transferred to higher trophic levels in the food web, trophic dilution
occurs.
• Overall, TCEP appears to be a persistent mobile organic compound (PMOC). PMOCs can
dissolve in water or bind to particles, resulting in longer environmental half4ives and greater
potential for long-range transport—especially in the air, water, and sediment compartments.
2.2.1 Fate and Transport Approach and Methodology
Reasonably available environmental fate data—including biotic and abiotic biodegradation rates,
removal during wastewater treatment, volatilization from lakes and rivers, and logarithmic organic
carbon:water partition coefficient (log Koc)—are the parameters used in the current risk evaluation. In
assessing the environmental fate and transport of TCEP, EPA considered the full range of results from
data sources that were rated high-quality. Information on the full extracted dataset is available in the
Page 42 of 638
-------
supplemental file RiskEvaluation for Tris (2-chloroethyl) Phosphate (TCEP) - Systematic Review of
Data Quality Evaluation and Data Extraction Information for Environmental Fate and Transport (U.S.
EPA. 2024t). Other fate estimates were based on modeling results from Estimation Programs Interface
(EPI) Suite™ (U.S. EPA. 2017a). a predictive tool for physical and chemical properties and
environmental fate estimation.3 Information regarding the model inputs is available in Appendix F.
Table 2-2 provides selected environmental fate data that EPA considered while assessing the fate of
TCEP and were updated after publication of Final Scope of the Risk Evaluation for Tris(2-chloroethyl)
Phosphate (TCEP) CASRN115-96-8 (U.S. EPA. 2020b) with additional information identified through
the systematic review process.
Table 2-2. Environmental Fate Properties of TCEP
Property or
Endpoint
Value"
Reference
Overall Quality
Determination
Indirect
photodegradation
ti/2 = 5.8 hours (based on -OH rate constant of
2.2E-11 cm3/mole-sec at 25 °C and 12-hour day
with 1.5E06 -OH/cm3; estimated) b
U.S. EPA (2017a)
High
Direct
photodegradation
Not expected to be susceptible to direct photolysis
by sunlight because the chemical structure of TCEP
does not contain chromophores that absorb at
wavelengths >290 nm
HSDB (2015)
High
Hydrolysis half-
life
ti/2 = 2 years at pH 8 and 25 °C (estimated)
Saint-Hilaire et al.
(2011)
High
ti/2 = 0.083 days at pH 13; no significant
degradation observed over 35 days at pH 7, 9, and
11
Su et al. (2016)
Aerobic
biodegradation
Water: 13% and 4%/28 days (OECD 301B) at 10
and 20 mg/L test substance concentration in
activated domestic sludge, adaption not specified
Life Sciences
Research Ltd
(1990b)
High
Soil: DT50 = 17.7 days; 78%/40 days based on test
substance concentration of 50 j^ig/kg
Hurtado et al.
(2017)
Anaerobic
biodegradation
No degradation observed
Pans et al. (2018)
High
Bioconcentration
factor (BCF) (L/kg,
unless noted)
Whole body BCF = 0.31 ± 0.06, 0.16 ± 0.03, and
0.34 ± 0.04 at test substance concentrations of 0.04,
0.2, and 1.0 mg/L, respectively in the muscle of
juvenile Atlantic salmon (Salmo scdar)
Arukwe et al.
(2018)
High
BCF = 1.0 ± 0.1 (muscle), 4.3 ± 0.2 (liver), 2.6 ±
0.2 (brain), 1.6 ± 0.1 (gill), and 1.6 ± 0.1 (kidney)
at test substance concentration of 9.1 (ig/L for
juvenile common carp (Cyprinus carpio) (OECD
305)
Tana et al. (2019)
BCF = 0.8 ± 0.1 (muscle), 2.4 ± 0.1 (liver), 2.2 ±
0.1 (brain), 1.9 ± 0.2 (gill) at test substance
concentration of 893 (ig/L, respectively for
zebrafish (Danio rerio) (OECD 305)
Wane et al. (2017a)
3 See EPI Suite™ for additional information and supporting documents about this freely available, online suite of programs,
which was reviewed by the EPA Science Advisory Board (SAB. 2007).
Page 43 of 638
-------
Property or
Endpoint
Value"
Reference
Overall Quality
Determination
Bioaccumulation
factor (BAF) (L/kg,
unless noted)
Mean BAF = 794 (muscle), 1,995 (liver), 1,995
(kidney), and 1,995 (gill)
Bekele et al. (2021)
High
Mean BAF = 30.7 (muscle) and 70.7 (liver) for
crucian carp {Caras sins auratus)
Choo et al. (2018)
Mean BAF = 2,198 at test substance concentration
of 0.464 ng/L for walleye (Sander vitreus)
Guo et al. (2017b)
Mean BAF = 1,248 for snakehead (Ophiocephalus
argus), 191 for catfish (Clarias batrachus), 109—
202 for mud carp (Cirrhinus molitorella), 207 for
crucian carp (Carassius auratus), and 463 for
Oriental River prawn (Macrobrachium nipponense)
Liu et al. (2019a)
Mean BAF = 6,310 for benthic invertebrates (soft
tissue); 2,690 for pelagic fish (organ); 4,270 for
benthic fish (organ and whole body)
Wans et al. (2019b)
Logarithmic
organic
carbon: water
partition coefficient
(log Koc) (soil)
2.08-2.52
Cristale et al.
(2017)
High
Logarithmic
organic
carbon: water
partition coefficient
(log Koc)
(sediment)
3.23 ±0.23
Wans et al.
(2018a)
High
3.32 (mean; range 2.5-4.06)
Zhans et al.
(2021)
3.46 ±0.48
Zhang et al.
(2018b)
Removal in
wastewater
treatment
Approximately -5% removal after primary
treatment; -19.1% overall removal
Kim et al. (2017)
High
Trophic
magnification
factor (TMF)
Benthic food web: 2.6 (tentative due to small
sample size, n = 15)
Brandsma et al.
High
No significant relationship with pelagic food web
and total food web
(2015)
Antarctic food chain: 5.2
Fu et al. (2020)
No significant relationship with trophic level
Zhao et al. (2018)
Biota-sediment
accumulation factor
(BSAF)
Mean BSAF (L/kg): 1.09 (muscle) and 2.49 (liver)
for Crucian carp (Carassius auratus)
Choo et al. (2018)
High
Mean BSAF: 0.015-0.171
Liu et al. (2019a)
Mean BSAF: 2.19E-03 for benthic invertebrates
and 1.48E-03 for benthic fishes
Wana et al. (2019b)
11 Measured unless otherwise noted
h Information estimated using EPI Suite™ (U.S. EPA. 2017a)
2.2.2 Summary of Fate and Transport Assessment
Numerous studies have described TCEP as a "ubiquitous" contaminant because it is commonly found in
various environmental compartments such as indoor air and dust, outdoor air, surface water, drinking
Page 44 of 638
-------
water, groundwater, soil, sediment, biota, and even precipitation all over the world (Awonaike et al..
2021; Ma et al.. 2021; Propp et al.. 2021; Choo and Oh. 2020; Li et al.. 2019b; Tan et al.. 2019; Arukwe
et al.. 2018; Kim and Kannan. 2018; Cao et al.. 2017; Hurtado et al.. 2017; Wang et al.. 2017a; Bradman
et al.. 2014; Padhye et al.. 2014; Cristate et al.. 2013; Bradman et al.. 2012; Regnery and Puttmann.
2010b; Benotti et al.. 2009; Fries and Puttmann. 2003. 2001). This is because TCEP is primarily used as
an additive plasticizer and flame retardant. When used as an additive, TCEP is added to manufactured
materials via physical mixing rather than chemical bonding and as a result, TCEP can easily leach or
diffuse into its surrounding environment (Oi et al.. 2019; Liu et al.. 2014; Wei et al.. 2014; AT SDR.
2012; van der Veen and de Boer. 2012; EC. 2009; ECB. 2009; NICNAS. 2001). TCEP's physical and
chemical properties suggests that its main mode of distribution in the environment is through water and
soil, depending on where it is being released (Figure 2-1; see also Appendix F.2.1.2) (U.S. EPA. 2017a;
TERA. 2015; Regnery and Puttmann. 2010b; Zhang et al.. 2009).
Multiple studies have identified urban sources as sources of TCEP in the environment through fugitive
emissions to air (Abdollahi et al.. 2017; Luo et al.. 2015; Moller et al.. 2011). The exact sources of
TCEP emissions from urban environment are unknown, however they are likely the articles that were
treated with or containing TCEP (Abdollahi et al.. 2017; Luo et al.. 2015; Wei et al.. 2014; Moller et al..
2011; Aston et al.. 1996). Compared to outdoor air, TCEP concentrations are significantly higher in
indoor air, because TCEP has the potential to volatilize from treated products and diffuse into air, as
well as partition onto dust, due to its use as an additive (Qi et al.. 2019; TERA. 2015; Liu et al.. 2014;
ATSDR. 2012; EC. 2009; NICNAS. 2001). Atmospheric deposition has been identified as an important
source of TCEP to surface water and soil, especially in urban areas. Several studies showed that higher
TCEP concentrations in precipitation were generally seen in densely populated areas with high traffic
volume (Kim and Kannan. 2018; Regnery and Puttmann. 2010b; Regnery and Puettmann. 2009;
Marklund et al.. 2005b). In addition, storm water and urban runoff can contribute to additional emissions
to surface water.
TCEP can be transported to sediment from overlying surface water by advection and dispersion of
dissolved TCEP and by deposition of suspended solids containing TCEP. However, TCEP may partition
between surface water and sediments to varying degrees because of its log Koc values (Zhang et al..
2021; Wang et al.. 2018a; Zhang et al.. 2018b) and water solubility (Lee et al.. 2018; Ma et al.. 2017;
Brandsma et al.. 2015; Cao et al.. 2012). which could contribute to its mobility in the environment.
Higher concentrations of TCEP in sediment are expected to be found at potential source locations (e.g.,
near urban and industrialized areas) (Chokwe and Okonkwo. 2019; Tan et al.. 2019; Lee et al.. 2018;
Wang et al.. 2018a; Cao et al.. 2017; Maruva et al.. 2016; Cristate et al.. 2013). Precipitation events,
such as rain and snow may enhance soil concentrations of TCEP, but accumulation in soil is expected to
be unlikely. TCEP may either volatilize from moist soil or migrate through the soil zone (Mihailovic et
al.. 2011). Due to its water solubility (7,820 mg/L) and soil log Koc values (2.08-2.52), dissolved TCEP
was observed to be mobile and migrated to groundwater by common soil transport processes such as
advection and diffusion (Propp et al.. 2021; Buszka et al.. 2009; Barnes et al.. 2004). TCEP in the soil
was seen to be vertically transported to deeper soil horizons, causing TCEP concentration in the surface
soil to be lower (He et al.. 2017; Bacaloni et al.. 2008).
Page 45 of 638
-------
Dry Deposition
Emissions
from
Source
Dispersal
Wet Deposition
indirect Atmospheric
Photolysis
tin= 5 8 h
Vapor pressure
0.0613 mmHg
Henry's Law Constant:
2.945E-06 atm-m3/mol
Wastewater facility
Indirect/Direct discharge
Aerobic Biodegradation
Rate = moderate
Indoor Air and Dust
Landfills
Surface Water
^ Aqueous
Photolysis
No degradation
Runoff
Land applied biosolids
Aerobic
Biodegradation
Rate = low
Anaerobic
Biodegradation
Rate = Negligible
Bioconcentration
BCF = 0,16-4 3
Hydrolysis =
Negligible
Sediment/
Pore Water
Groundwater
Anaerobic
Biodegradation
Rate = Negligible
Figure Legend
Negligible
¦
Partitioning/T ransportation
~
T rarisformation/Degradation
¦
Wastewater Facility
Figure 2-1. Transport, Partitioning, and Degradation of TCEP in the Environment"
* The diagram depicts the distribution (grey arrows), transport and partitioning (black arrows), and the
transformation and degradation (white arrows) of TCEP in the environment. The width of the arrow is a
qualitative indication of the likelihood that the indicated partitioning will occur or the rate at which the indicated
degradation will occur (i.e., wider arrows indicate more likely partitioning or more rapid degradation).
Most flame retardants that have "High" or "Very High" persistence designations, such as TCEP, are
persistent because they are expected to be stable by design to maintain their flame-retardant properties
throughout its lifetime in products ( J.S. EPA. 2015a). Based on multiple monitoring studies, TCEP
appears to be a persistent mobile organic compound (PMOC). PMOCs can dissolve in water or bind to
particles, resulting in longer environmental half-lives and greater potential for long-range transport
(Blum et al . 2019; Rodgers et al.. 2018; Reemtsma et al.. 2016). TCEP was detected in both lake and
Page 46 of 638
-------
marine waters of the Arctic, where TCEP was quantified in water and air far from human settlements
(>500 km). Atmospheric deposition and watershed runoff may be the primary sources of TCEP in these
remote waters where TCEP is unlikely to be rapidly transformed by hydrolysis, photolysis, or
biodegradation (Na et al.. 2020; McDonough et al.. 2018; Li et al.. 2017b). These findings indicate that
TCEP has the potential to undergo long-range transport in air and water. TCEP's long-range transport
potential (LRTP) was seen to be significantly underestimated when using its physical and chemical
properties in quantitative structure-activity relationship (QSAR) models because the behavior of TCEP
in the environment often does not align with its physical and chemical properties. A detailed summary
of physical and chemical properties and a fate and transport assessment of TCEP is available in
Appendix F.
2.2.3 Weight of Scientific Evidence Conclusions for Fate and Transport
2.2.3.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the
Fate and Transport Assessment
Given the consistent results from numerous high-quality studies, there is a robust confidence that TCEP:
• is not expected to undergo significant direct photolysis (see Appendix F.2.2);
• will partition to organic carbon in the air (see Appendix F.2.2);
• will exist in both the gas and particle phases (see Appendix F.2.2);
• showed no significant degradation after undergoing hydrolysis but hydrolysis rate was seen to
increase with increasing pH (see Appendix F.2.3.1);
• does not undergo biodegradation in water under aerobic conditions (see Appendix F.2.3.1);
• will volatilize from surface water (see Appendix F.2.3.1) and moist soil (see Appendix F.2.4.1);
• produces hazardous byproducts when undergoing thermal degradation (see Appendix F.2.5.1);
• will not be removed after undergoing wastewater treatment and will be retained in effluents with
low fraction being adsorbed onto sludge (see Appendix F.2.5.2);
• is minimally removed after undergoing conventional drinking water treatment (see Appendix
F.2.5.3); and
• has the ability to undergo long-range transport (see Appendices F.2.2 and F.2.3.1).
As a result of limited studies identified, there is a moderate confidence that TCEP:
• does not undergo biodegradation in water and sediment under anaerobic conditions (see
Appendixes F.2.3.1 and F.2.3.2);
• will partition to organic carbon in sediment (see Appendix F.2.3.2) and soil (see Appendix
F.2.4.1);
• will enter surface water and groundwater from landfills (see Appendix F.2.4.3);
• will not bioaccumulate in fish residing in the water column (see Appendix F.2.6);
• may bioaccumulate in benthic fish (see Appendix F.2.6); and
• does not bioaccumulate when TCEP concentrations are transferred to higher trophic levels in the
food web (see Appendix F.2.6).
A detailed discussion of strengths, limitations, assumptions, and key sources of uncertainty for the fate
and transport assessment of TCEP is available in Appendix F.
Page 47 of 638
-------
3 RELEASES AND CONCENTRATIONS IN THE ENVIRONMENT
EPA estimated environmental releases of TCEP. Section 3.1 describes the approach and methodology
for estimating releases. Estimates of environmental releases are presented in Section 3.2. Section 3.3
presents the approach, methodology, and estimates of environmental concentrations that result from
environmental releases of TCEP.
3.1 Approach and Methodology
3.1.1 Industrial and Commercial
EPA categorized the COUs listed in Table 1-1 into occupational exposure scenarios (OESs) (see Table
3-1). EPA developed the OESs to group processes or applications with similar sources of release and
occupational exposures that occur at industrial and commercial workplaces within the scope of the risk
evaluation. For each OES, occupational exposure and environmental release results are provided and
expected to be representative of the entire population of workers and sites involved for the given OES in
the United States. Note that EPA may define only a single OES for multiple COUs, while in other cases
multiple OESs may be developed for a single COU. For example, the paint and coating manufacturing
COU has two associated OESs—a 1-part coatings scenario and a 2-part reactive coatings scenario. EPA
makes this determination by considering variability in release and use conditions and whether the
variability can be captured as a distribution of exposure or instead requires discrete scenarios.
Specifically, the 1-part coatings tend to be water-based formulations and could potentially have a greater
release to water whereas the 2-part reactive coatings could have greater release to incineration or
landfill. Further information on specific OESs is provided in Supplemental Information on
Environmental Release and Occupational Exposure Assessment (U.S. EPA 2024n).
All COUs assessed in this Risk Evaluation are considered on-going uses. However, there are several
COUs for which part of the life cycle has ceased, such as manufacturing (including import) and
processing. However, other parts of the life cycle including recycling, commercial or consumer use, and
disposal are on-going. These COUs are identified in Table 3-1 and include four COUs for commercial
use and five COUs for consumer use.
Page 48 of 638
-------
Table 3-1. Crosswalk of COUs to OESs Assessed
cou
OES
Life Cycle Stage"
Category6
Subcategoryc
Manufacture
Import
Import
Repackaging
Processing
Incorporated into
formulation, mixture, or
reaction product
Paint and coating
manufacturing
Incorporation into paints and
coatings - 1-part coatings
Incorporation into paints and
coatings - 2-part reactive coatings
Incorporated into
formulation, mixture, or
reaction product
Polymers used in aerospace
equipment and products
Formulation of TCEP into 2-part
reactive resins
Incorporated into article
Aerospace equipment and
products and automotive
articles and replacement
parts containing TCEP
Processing into 2-part resin article
Recycling
Recycling
Recycling e-waste
Distribution
Distribution
Distribution in commerce
Distribution in commerce'
Industrial Use
Other use
Aerospace equipment and
products and automotive
articles and replacement
parts containing TCEP
Installation of article
Paints and coatings
Paints and coatings
Use of paints and coatings - Spray
application OES
Commercial Use
Other use
Aerospace equipment and
products and automotive
articles and replacement
parts containing TCEP
Use and/or maintenance of
aerospace equipment and products
and automotive articles and
replacement parts containing TCEP
Paints and coatings
Paints and coatings
Use of paints and coatings - Spray
application OES
Other use
Laboratory chemicals
Lab chemical - Use of laboratory
chemicals
Furnishing, cleaning,
treatment/care products
Fabric and textile products''
End of service life disposal1'
(releases and exposures not
quantified)
Foam Seating and Bedding
Products1'
End of service life disposal1'
(releases and exposures not
quantified)
Construction, paint,
electrical, and metal
products
Building/construction
materials - Insulation''
End of service life disposal1'
(releases and exposures not
quantified)
Building/construction
materials - Wood and
engineered wood products -
Wood resin composites''
End of service life disposal1'
(releases and exposures not
quantified)
Page 49 of 638
-------
COU
OES
Life Cycle Stage"
Category6
Subcategoryc
Disposal
Disposal
Disposal6
Waste disposal (landfill or
incineration, covered in each
COU/OES as opposed to a separate
COU)
° Life Cycle Stage Use Definitions (40 CFR 711.3)
- "Industrial Use" means use at a site at which one or more chemicals or mixtures are manufactured (including
imported) or processed.
- "Commercial Use" means the use of a chemical or a mixture containing a chemical (including as part of an article) in
a commercial enterprise providing saleable goods or services.
- "Consumer Use" means the use of a chemical or a mixture containing a chemical (including as part of an article, such
as furniture or clothing) when sold to or made available to consumers for their use.
- Although EPA has identified both industrial and commercial uses here for purposes of distinguishing scenarios in this
document, the Agency interprets the authority over "any manner or method of commercial use" under TSCA section
6(a)(5) to reach both.
b These categories of COUs appear in the LCD, reflect CDR codes, and broadly represent COUs of TCEP in industrial
and/or commercial settings and for consumer uses.
c These subcategories reflect more specific COUs of TCEP.
d This COU includes associated disposal of those COUs for which domestic manufacturing and/or processing have ceased.
e Section 3.2 provides details on these OESs.
^Distribution in Commerce of TCEP consists of the transportation associated with the moving of sealed containers of
TCEP or sealed packages of TCEP containing products.
The 2016 CDR data (U.S. EPA. 2019a) included a single reporting site, Aceto Corporation in Port
Washington, New York, importing TCEP, with no downstream industry sectors identified. TCEP was
not reported in the 2020 CDR (U.S. EPA. 2020a). EPA did identify other data on current import
volumes and possible import sites from Datamyne, as presented in Figure 1-3, which showed some
TCEP imports below the CDR threshold of 25,000 lb/site-yr. Nevertheless, processors of TCEP may be
purchasing the chemical from importers (see Supplemental Information on Environmental Release and
Occupational Exposure Assessment (U.S. EPA. 2024n) for details). Therefore, EPA assumed TCEP may
still be imported at volumes below the CDR reporting threshold and EPA assessed the following two
potential scenarios: (1) one site importing 25,000 lb, and (2) one site importing 2,500 lb. EPA modeled
environmental releases and occupational exposures for these hypothetical scenarios. For each OES,
where monitoring data were not available, daily releases were estimated per media of release based on
EPA Standard Models, Generic Scenarios (GSs), and/or Emission Scenario Documents (ESDs) to
generate annual releases and for the estimation of associated release days. TCEP is not listed on the
National Emissions Inventory (NEI) and was only recently added to TRI, with the first year of reporting
from facilities due July 1, 2024. As of September 2024, there are no TRI entries for TCEP. EPA
describes its approach and methodology for estimating daily releases and for detailed facility level
results in Supplemental Information on Environmental Release and Occupational Exposure Assessment
(U.S. EPA. 2024n).
Page 50 of 638
-------
Monitoring
Data
HSIA, Reports,
NIOSH, OSHA
Inhalation
Exposure
Modeling
NF/FF, ESD
DEVL model
Modeling
Dermal
Exposure
OES
Occupational
Assessment
# Workers or
ONUs per site
BLS, Census,
ESD
# of Workers,
ONUs Exposed
Number of
facilities
Census, NEI,
TRI, DMR, CDR
Figure 3-1. An Overview of How EPA Estimated Daily Releases for Each OES
BLS = Bureau of Labor Statistics; DEVL = Dermal Exposure to Volatile Liquids model; DMR = Discharge
Monitoring Report; ELG = Effluent Limitation Guidelines; HSIA = Halogenated Solvents Industry Alliance;
NF/FF = Near-Field/Far Field; NIOSH = National Institute for Occupational Safety and Health; OSHA =
Occupational Safety and Health Administration
The releases of TCEP were estimated for each media applicable to the OES. For TCEP, releases could
occur to water, air, or disposal to land. TCEP released could be in the form of liquid (neat or in
formulation), vapor, and/or solid waste.
Page 51 of 638
-------
3.2 Environmental Releases
Environmental Releases (Section 3.2):
Key Points
EPA evaluated the reasonably available information for releases of TCEP to the environment. The
key points of the environmental releases are summarized below:
• EPA assessed environmental releases of TCEP from industrial and commercial sources as
well as consumer products.
o For industrial and commercial sources, EPA used data from literature, relevant ESDs, or
GSs to estimate environmental releases to air, surface water, and waste disposal from a
generic facility for each OES. Some OESs could not be quantified due to insufficient
data. Of the OESs that could be quantified, the highest release estimates were from:
¦ Incorporation into paints and coatings - 1-part coatings
¦ Incorporation into paints and coatings - 2-part reactive coatings
¦ Formulation of TCEP-containing reactive resins (for use in 2-part systems)
¦ Use of paints and coatings - spray application OES.
o For consumer products, EPA did not have enough information to assess environmental
releases quantitatively. However, the Agency acknowledges that there may be TCEP
releases to the environment via the demolition and disposal of consumer articles, as well
as to wastewater via domestic laundry. These releases were assessed qualitatively. EPA
included anecdotal information from peer-reviewed literature on releases from consumer
articles in Section 5.1.2.2.5.
3.2.1 Industrial and Commercial
EPA combined its estimates for each activity that is reasonably expected to occur during each OES.
These activities were based on using data from literature, relevant ESDs or GSs. Once these activities
were identified, existing EPA models and parameters (e.g., the EPA/OPPT Mass Transfer Coefficient
model, EPA/OPPT Penetration model, ChemSTEER User Guide, etc.) were used in a Monte Carlo
simulation to create a distribution of releases. From this distribution EPA provides a high-end (95th
percentile) and central tendency (50th percentile) release values as well as a range of potential release
days. The releases presented are assumed to be representative of what would be reasonably expected to
occur at an individual generic site. In some cases, where it was not reasonable to assume a single generic
site due to throughput constrictions presented in the relevant source (e.g., it is not reasonable to assume
that a single paint application site or laboratory would use the entire PV of 25,000 lb), a range of
potential number of sites is presented in Table 5-2. A summary of these ranges of releases across OESs
is presented in Table 3-2. See Supplemental Information on Environmental Release and Occupational
Exposure Assessment (U.S. EPA. 2024n) for more details on deriving the overall confidence score for
each OES. For some OESs, EPA was not able to estimate or did not anticipate there to be releases; for
example:
• EPA was not able to quantify disposal of articles that historically contained TCEP with
reasonably available information. This was assessed qualitatively.
• Installation of articles is not expected to lead to significant releases because the articles are
expected to already be in final form (e.g., electronic potting) and not expected to undergo further
processing (i.e., shaping, sanding, cutting, etc.).
Page 52 of 638
-------
• EPA was not able to quantify releases of TCEP that could occur during the recycling of
electronic waste (e-waste). Sources used for this provided monitoring data from breathing zone
measurements from various locations within a facility that recycles e-waste that contained very
small amounts of TCEP dust. The source of TCEP was not identified and the source further
stated that TCEP is rarely used in electronics. EPA expects releases that could occur during this
activity to be minimal and only potentially occur at a small subset of facilities.
• EPA lacks production volume data to assess TCEP exposure from distribution into commerce
due to the declining production and manufacturing in recent years. Although manufacturing,
processing, and distribution into commerce of TCEP is declining (see Section 1.1.1, Table 3-1);
distribution into commerce that has occurred, is ongoing, or is likely to occur during a COU
subject to evaluation; and exposure to human or ecological populations has occurred or is likely
to occur; will be included in the risk evaluation as an exposure associated with a COU.
3.2.1.1 Summary of Daily Environmental Release Estimates
Table 3-2 and Table 3-3 provide estimated releases that could occur during each OES, the expected
media of release if releases are expected to occur during that OES, and possible number of sites where
releases could occur. The estimated daily releases are based on a 2,500 lb production volume. For most
cases, the number of sites is based on a single generic site; however, in some cases, such as use of paints
and coatings and laboratory chemicals, a distribution of the number of sites was created. The
distributions for number of sites were created for these OESs to provide variability in the potential
number of sites and is further explained in the Supplemental Information on Environmental Release and
Occupational Exposure Assessment (U.S. EPA 2024nV
Page 53 of 638
-------
Table 3-2. Summary of EPA's Daily Release Estimates for Each OES and EPA's Overall Confidence in these Estimates for 2,500 lb
Production Volume
cou
OES
Estimated Daily Release
Range across Sites
(kg/site-day)
Type of Discharge," Air
Emission,* or Transfer for
Disposal'
Estimated Release
Frequency Range
across Sites (days)''
Number of
Facilities^
Overall
Confidence
Central
Tendency
High-End
Central
Tendency
High-
End
Manufacture
(Import)
Repackaging
6.35
9.89
Surface water
4
4
1 generic
site
Medium
3.18E-04
6.03E-04
Fugitive or stack air
4
4
N/A
N/A
Waste disposal (landfill or
incineration)
N/A
N/A
Processing
Incorporation into paints and
coatings - 1-part coatings
1.02E01
3.52E01
Surface water
6
2
1 generic
site
High
1.56E-03
9.60E-03
Fugitive or stack air
6
4
1.53
9.27
Waste disposal (landfill or
incineration)
7
2
Processing
Incorporation into paints and
coatings - 2-part reactive coatings
2.71E01
3.19E01
Surface water
1
1
1 generic
site
High
3.65E-03
7.90E-03
Fugitive air
1
1
3.75E-03
1.99E-02
Stack air
1
1
3.40E01
3.40E01
Waste disposal (landfill or
incineration)
1
1
Processing
Formulation of TCEP-containing
reactive resins (for use in 2-part
systems)
2.52E01
3.15E01
Surface water
1
1
1 generic site
High
3.25E-03
8.83E-03
Fugitive air
1
1
2.73E-03
2.07E-02
Stack air
1
1
3.40E01
3.40E01
Waste disposal (landfill or
incineration)
1
1
Processing
Processing into 2-part resin article
N/A
N/A
Surface water
N/A
N/A
1 generic
site
High
3.30E-04
9.90E-04
Fugitive or stack air
55
113
3.98E-01
2.50
Waste disposal (landfill or
incineration)
92
17
Processing
Recycling e-waste
EPA did not have sufficient data to estimate these releases
Distribution
Distribution in commerce
Distribution in Commerce h
Industrial Use
Installation of articles
Releases expected to be negligible
Page 54 of 638
-------
cou
OES
Estimated Daily Release
Range across Sites
(kg/site-day)
Type of Discharge," Air
Emission,* or Transfer for
Estimated Release
Frequency Range
across Sites (days)''
Number of
Facilities^
Overall
Confidence
Central
Tendency
High-End
Disposal'
Central
Tendency
High-
End
2.37
2.32E01
Surface water
1
2
95th
Industrial and
Industrial and Commercial Use of
1.25E01
1.14E02
Fugitive air
1
2
Percentile:
2,031
50th
Percentile:
281
commercial
use
paints and coatings - Spray
application8
N/A
N/A
Waste disposal (landfill or
incineration)
N/A
N/A
Medium
Use and/or maintenance of
aerospace equipment and products
and automotive articles and
replacement parts containing TCEP
Releases expected to be negligible
3.96E-01'
8.83E-01'
Surface water
220
214
13 (1st
percentile) -
Lab chemical - Use of laboratory
6.47E-05^
7.99E-05^
Fugitive or stack air
220
214
High
chemicals
N/A
N/A
Waste disposal (landfill or
incineration)
N/A
N/A
6 (5th
percentile)
Commercial
Use
Furnishing, cleaning,
treatment/care products
• Fabric and textile products
• Foam seating and bedding
products
Construction, paint, electrical, and
metal products
• Building/construction materials
- Insulation
• Building/construction materials
- Wood and engineered wood
products - Wood resin
composites
Manufacturing and Processing of these COU's has ceased, EPA does not have sufficient data to estimate the
releases that may occur during disposal of already existing products
Page 55 of 638
-------
cou
OES
Estimated Daily Release
Range across Sites
(kg/site-day)
Type of Discharge," Air
Emission,* or Transfer for
Disposal'
Estimated Release
Frequency Range
across Sites (days)''
Number of
Facilities^
Overall
Confidence
Central ... , ,
„ . High-End
Tendency
Central High-
Tendency End
Disposal
Disposal
Waste Disposal (Landfill or Incineration, covered in each COU/OES as opposed to a separate COU)
" Direct discharge to surface water; indirect discharge to non-POTW; indirect discharge to POTW
h Emissions via fugitive air; stack air; or treatment via incineration
c Transfer to surface impoundment, land application, or landfills
d Where available, EPA used peer-reviewed literature (e.g., GSs or ESDs) to provide a basis to estimate the number of release days of TCEP within a COU.
e Where available, EPA used peer reviewed literature (e.g., emission scenario documents) data to provide a basis to estimate the number of sites using TCEP within a
condition of use.
^"High-end" is the 5th percentile and "Central Tendency" is the 1st percentile. See Section 3.10 of Engineering Supplemental file for rationale of using the 1st and 5th
percentiles.
g Multiple throughput and site scenarios are presented in Table 5-1 of the Engineering Supplemental file.
h Distribution in Commerce of TCEP consists of the transportation associated with the moving of sealed containers of TCEP or sealed packages of TCEP containing
products.
Page 56 of 638
-------
Table 3-3. Summary of EPA's Release Estimates for Each COl
J/OES and EPA's Overal
Confidence in these Estimates
Life Cycle
Stage
Surface
Water
Air
Waste Disposal
Overall
Confidence
Category
Subcategory
OES
Fugitive
Air
Stack
Air
Landfill
Incineration
Sources
Manufacture
(Import)
Import
Import
Repackaging
0
0
0
0
0
Medium
Peer-reviewed
literature6
(GS/ESD)
Incorporated
into
formulation,
Paint and coating
Incorporation into
paints and coatings -
1-part coatings
0
0
0
0
0
High
Peer-reviewed
literature6
(GS/ESD)
mixture, or
reaction
product
manufacturing
Incorporation into
paints and coatings -
2-part coatings
0
0
0
0
0
High
Peer-reviewed
literature6
(GS/ESD)
Incorporated
into
formulation,
mixture, or
reaction
product
Polymers used in
aerospace
equipment and
Formulation of
TCEP-containing
reactive resins (for
use in 2-part
systems)
0
0
0
0
0
High
Peer-reviewed
literature'
(GS/ESD)
Processing
Incorporated
into article
Aerospace
equipment and
products and
automotive articles
and replacement
parts containing
TCEP
Processing into 2-
part resin article
0
0
0
0
0
High
Peer-reviewed
literature6
(GS/ESD)
Recycling
Recycling
Recycling e-waste
~
~
~
~
~
Medium
NIOSH HHE's
used for
exposure
estimates;
insufficient
data to
estimate
releases
Page 57 of 638
-------
Life Cycle
Stage
Surface
Water
Air
Waste Disposal
Overall
Confidence
Category
Subcategory
OES
Fugitive
Air
Stack
Air
Landfill
Incineration
Sources
Distribution
Distribution
Distribution in
commerce
Distribution in
Commerce
Distribution (Table 3-1)
Industrial
Use
Other use
Aerospace
equipment and
products and
automotive articles
and replacement
parts containing
TCEP
Installation of article
0
0
0
0
0
Medium
Releases not
expected to
occur during
handling of
aerospace
articles
Paints and
coatings
Paints and coatings
Use of paints and
coatings - Spray
application OES
1,000 kg daily
throughput
0
0
0
0
0
Medium
Peer-reviewed
literature®
Page 58 of 638
-------
Life Cycle
Stage
Surface
Water
Air
Waste Disposal
Overall
Confidence
Category
Subcategory
OES
Fugitive
Air
Stack
Air
Landfill
Incineration
Sources
Other use
Aerospace
equipment and
products and
automotive articles
and replacement
parts containing
TCEP
Use and/or
maintenance of
aerospace equipment
and products
0
0
0
0
0
Medium
Releases not
expected to
occur during
handling of
aerospace
articles
Paints and
coatings
Paints and coatings
Use of paints and
coatings - Spray
application OES
1,000 kg daily
throughput
0
0
0
0
0
Medium
Peer-reviewed
literature6
Other use
Laboratory
chemicals
Lab chemical - Use
of laboratory
chemicals
0
0
0
0
0
Peer-reviewed
literature6
Commercial
Use
Furnishing,
cleaning,
Fabric and textile
products
~
~
~
~
~
Medium
Peer-reviewed
literature6
treatment/care
products
Foam seating and
bedding products
~
~
~
~
~
Medium
Peer-reviewed
literature6
Construction,
paint, electrical,
and metal
products
Building/
construction
materials -
Insulation
~
~
~
~
~
Medium
Peer-reviewed
literature6
Building/
construction
materials - Wood
and engineered
wood products -
Wood resin
composites
~
~
~
~
~
Medium
Peer-reviewed
literature6
Disposal
Disposal
Evaluated as part of each OES
as opposed to a standalone OES
0= Estimated releases 13= Estimated releases but not anticipated = Releases not quantified, assessed qualitatively
Page 59 of 638
-------
3.2.2 Consumer Releases
Environmental releases to the environment may occur from consumer articles containing TCEP via the
end-of-life disposal and demolition of consumer articles in the built environment, as well as from the
associated down-the-drain release of TCEP from domestic laundry that removes TCEP containing dust
from clothing to wastewater. It is difficult for EPA to quantify these ends-of-life and down-the-drain
laundry exposures due to limited reasonably available information on source attribution of the consumer
COUs. In previous assessments, EPA has considered down-the-drain analysis for consumer products
scenarios where there is reasonably foreseen exposure scenario where it can be assumed the consumer
product (e.g., drain cleaner, lubricant, oils) will be discarded directly down-the-drain. Although EPA
acknowledges that there may be TCEP releases to the environment via the demolition and disposal of
consumer articles and the release of TCEP to wastewater via domestic laundry, the Agency did not
quantitatively assess these scenarios due to lack of reasonably available information. EPA instead
assessed them qualitatively. Anecdotal information in the peer-reviewed literature helps qualitatively
describe how TCEP may be potentially released to the environment from consumer articles (see Section
5.1.2.2.5).
3.2.3 Weight of Scientific Evidence Conclusions for Environmental Releases from
Industrial, Commercial, and Consumer Sources
For each OES, EPA considered the assessment approach, the quality of the data and models, and
uncertainties in assessment results to determine a level of confidence as presented in Supplemental
Information on Environmental Release and Occupational Exposure Assessment (U.S. EPA 2024n).
EPA determined that the various GSs and ESDs had overall quality determinations of high or medium,
depending on the GS/ESD. The GSs and ESDs are documents developed by EPA or OECD that are
intended to provide an overview of an industry and identify potential release and exposure points for that
industry; they cover processes and are not specific to any chemical. This lack of chemical specificity
creates an uncertainty in the overall release estimate—the assessed parameter values may not always be
representative of applications specific to TCEP use in each OES. Another uncertainty is lack of
consideration for release controls. The GS/ESDs assume that all activities occur without any release
controls and in an open-system environment where vapor and particulates freely escape. Actual releases
may be less than estimated if facilities utilize pollution control methods. Although TCEP monitoring
data would be preferred to modeled estimates from generic scenarios, monitoring data were not
available for almost all the OESs included in the risk evaluation. EPA strengthened modeled estimates
by using Monte Carlo modeling to allow for variation in environmental release calculation input
parameters according to the GS/ESD and other literature sources. The Agency was unable to
quantitatively assess releases to the environment from consumer products containing TCEP. See section
5.3.2.3.2 for additional details.
3.2.3.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the
Environmental Release Assessment
Use of Reporting Year-Release Trends Analysis
The 2016 CDR only had one reporter of TCEP while the 2020 CDR had no reporters; it is assumed that
TCEP has been largely phased out of products it was historically used in such as flexible and rigid foam
products. EPA expects that any remaining current users of TCEP do not surpass the CDR reporting
threshold of 25,000 lb per site-year (i.e., <25,000 lb/year is used at any given site).
EPA searched the Discharge Monitoring Report (DMR) database for TCEP monitoring data from 2010
to 2021. Monitoring data were available for locations in California; however, TCEP was not detected in
any of the effluents of the POTWs that were monitored (U.S. EPA 2022a). DMR data are submitted by
Page 60 of 638
-------
NPDES permit holders to states or directly to EPA according to the monitoring requirements of the
facility's permit. States are required to load only major discharger data into DMR and may or may not
load minor discharger data. The definition of major vs. minor discharger is set by each state and could
be based on discharge volume or facility size. Due to these limitations, some sites that discharge may
not be included in the DMR dataset. It is uncertain the extent to which sites not captured in these
databases release TCEP into the environment or whether the releases are to water, air, or landfill. TCEP
was officially added to TRI at the end of 2022. However, companies will not have to report on their
possible management and/or use of TCEP until July 2024. As of September 2024, there are no TRI
entries for TCEP.
EPA also searched other databases including the Water Quality Portal (WQP), where monitoring trends
indicate a downward trend of TCEP concentrations in surface water (see Figure 3-9).
Use of Generic Scenario and Emission Scenario Documents for Number of Facilities
In some cases, the number of facilities for a given OES was estimated using GSs and ESDs, which are
peer-reviewed. These documents typically attempt to find and map applicable North American Industry
Classification System (NAICS) codes to an OES. This is done by identifying keywords relevant to that
OES and entering them into the search tool on the U.S. Census Bureau's website. The results are
reviewed for relevancy and the most applicable NAICS codes are selected. It is possible that the NAICS
codes selected may not fully represent all potential types of sites for a given OES.
Uncertainties Associated with Number of Release Days Estimate
EPA did not have site specific data for the number of release days for most OESs. Typically, in these
cases, the Agency assumed that an activity occurs once per day (e.g., a facility may process a single
batch per day). In the event that this assumption leads to a number of operating days that exceeds 365
days, it may be assumed that a site will be processing more than one batch per day. Given the relatively
small production volume of TCEP being assessed this situation was not encountered. However, it is
possible that this could lead to either an under or over estimation of the number of release days. In
certain circumstances, EPA chose 250 days a year as the upper bound of possible number of operating
days because that is considered the maximum number of days a worker would be exposed, but for most
OESs the number of release days was well under this value.
Page 61 of 638
-------
Concentrations in the Environment (Section 3.3):
Key Points
EPA evaluated the reasonably available information on concentrations of TCEP in the environment.
The key points on environmental concentrations are summarized below:
• EPA assessed environmental concentrations of TCEP in air, water, and land (soil, biosolids
and groundwater).
o For the air pathway, measured data from a variety of locations within and outside of the
United States provided TCEP concentrations near facilities and locations that would
represent general population exposure, as well as in remote locations. EPA also modeled
ambient air concentrations and deposition from facilities releasing TCEP to air. The
Agency expects dry and wet air deposition of TCEP from air to land and surface waters
may be an important source of TCEP to the ambient environment.
o For the water pathway, EPA found measured data on TCEP in surface water,
precipitation, groundwater, wastewater, and the sediment compartment. The Agency also
modeled TCEP concentrations in surface water and sediment, including air deposition
contributions to each, near facilities releasing TCEP. EPA expects surface water and
sediment to be the main environmental exposure pathways for aquatic organisms.
o For the land pathway, EPA found measured concentrations of TCEP in soil, biosolids,
and groundwater. The Agency modeled soil concentrations from air deposition and
biosolids as well as groundwater concentrations from landfill leachate. EPA does not
expect TCEP concentrations to accumulate in soil; rather, TCEP in soil is expected to
migrate to groundwater.
3.3 Concentrations in the Environment
The environmental exposure characterization focuses on aquatic and terrestrial releases of TCEP from
hypothetical facilities that use, manufacture, or process TCEP under industrial and/or commercial COUs
subject to TSCA regulations. To characterize environmental exposure, EPA assessed point estimate
exposures derived from both measured and predicted concentrations of TCEP in ambient air, surface
water, and landfills in the United States.
A literature search was also conducted to identify sources of TCEP monitoring and reported modeled
data. The tornado plots in the subsequent sections are a summary of the monitoring for the various
environmental media. The plots provide the range of media concentrations in monitoring various
studies. The plots are split by U.S. and non-U.S. data, fraction (e.g., vapor, gas, particle; see Figure 3-2
and the studies are ordered from top to bottom from newer to older data. The plots are colored to
indicate general population, remote, near facility, and unknown population information.
For more information on TCEP monitoring data, please see the following documents:
• Environmental Monitoring Concentrations Reported by Media Type (U.S. EPA. 2024i);
• Environmental Monitoring and Biomonitoring Concentrations Summary Table (U.S. EPA.
2024h);
• Data Quality Evaluation Information for General Population, Consumer, and Environmental
Exposure. (U.S. EPA. 2024x); and
• Data Extraction Information for General Population, Consumer, and Environmental Exposure
(U.S. EPA 2024r).
Page 62 of 638
-------
3.3.1 Ambient Air Pathway
EPA conducted systematic review to obtain concentrations of TCEP in ambient air. Section 3.3.1.1
displays the aggregated results of reported monitoring concentrations for ambient air found from the
systematic review. Section 3.3.1.2 reports EPA modeled ambient air concentrations and deposition
fluxes.
Ambient air concentrations of TCEP were measured in seven studies in the United States (Figure 3-2).
Bradman et al. (2014) recorded a maximum concentration of 1.60 ng/m3 at 14 early childhood education
facilities in California between May 2010 and May 2011. Peverly et al. (2015) sampled TCEP in
ambient air at 13 locations across Chicago, Illinois. They demonstrated that TCEP ambient air
concentrations (maximum of 0.570 ng/m3) were slightly higher nearer to downtown Chicago than
suburban Chicago.
Moran et al. (2023) detected atmospheric concentrations of TCEP at Troutman Lake, AK of up to 2.8
ng/m3. The air concentrations were in the vapor phase. TCEP was also found to be in deposition to
Troutman Lake at magnitudes of 290 to 1,300 ng/m2/day (Moran et al.. 2023). Although TCEP potential
for LRTP has been described in the literature, other comparable Arctic sites suggest local point sources
may contribute to the vapor phase TCEP concentrations around Troutman Lake. Furthermore, passive
samplers deployed near the Native Village of Savoonga, 63 km from Troutman Lake, reported TCEP
below the detection limit of (0.01 ng/m3), one to two orders of magnitude lower than the concentrations
reported by (Moran et al.. 2023). Troutman lake lies directly south of the Village of Gambell on the
northwest corner of Sivuqaq. The island is home to the Sivuqaq Yupik people who practice a traditional
subsistence lifestyle. During the Cold War, a military instillation was developed on Sivuqaq due to its
proximity to Russia. Upon closure of the installation, debris and chemical waste were buried by the
military in disposal sites around the village and the adjacent Troutman Lake. Moran et al. (2023)
suggests that the occurrence of these organophosphate flame retardants is unlikely to be associated with
formerly used defense sites, because the defense site precedes the use of these chemicals. Rather, Moran
et al. (2023) suggests that local waste disposal practices and nearby open-air landfills may represent
important sources of TCEP concentrations to Troutman Lake. Although the Gambell West landfill was
900 m and 1,400 m from the northwest and northeast sampling locations, respectively, the concentration
reported in this study (2.8 ng/m3) was similar to those reported in Kerric et al. (2021). which described
mean combined vapor and gas concentrations of 2.4 ng/m3 around a municipal landfill near Montreal,
Canada. The Gambell West Landfill is a class III community active trench and fill landfill that uses burn
units for treatment. Deposition of TCEP was higher on the north end of the lake than the south end of
the lake. The northwest site had the most deposition with a magnitude of 1,300 ng/m2/day.
Page 63 of 638
-------
3.3.1.1 Measured Concentrations in Ambient Air
g General Population (Background)
¦ Remote (Not Near Source)
¦¦ Near Facility < Highly Exposed)
¦ Unknown/Not Specilied
V Lognonnal Distribution (CT and 90th percentile)
A Normal Distribution
-------
Industrial and commercial release estimates are presented in Section 3.2. Table 3-3 provides the
following COUs/OESs that have ambient air releases (stack or fugitive). These facility releases were
utilized to model ambient air concentrations and deposition via AERMOD and IIOAC.
The full set of inputs and results of IIOAC and AERMOD are presented in Appendix 1.3. For the initial
IIOAC runs, EPA modeled each of the fugitive air and stack air release scenarios for the seven relevant
OESs. In addition, due to initial uncertainties in the particle size, EPA ran IIOAC for both fine and
coarse particle settings for TCEP. In IIOAC, all calculated air concentrations of fine and coarse particles
are capped by an upper limit equal to the National Ambient Air Quality Standards (NAAQS) for
particulate matter (PM). These limits are 35 and 150 [j,g/m3 for fine and coarse particles (i.e., the
NAAQS for PM2.5 and PM10), respectively. These limits were met for all the OESs with stack
emissions. In addition, this limit was reached for the fine particle size, fugitive emissions run for the
commercial use of paints and coatings (see Appendix 1.3).
A further limitation of IIOAC is that it does not model gaseous deposition. Due to the inability to model
gaseous deposition, and due to the initial screening results meeting the NAAQS caps, EPA decided to
run a higher tier model (AERMOD) for the ambient air pathway.
AERMOD is a steady-state Gaussian plume dispersion model that incorporates air dispersion based on
planetary boundary layer turbulence structure and scaling concepts, including treatment of both surface
and elevated sources and both simple and complex terrain. AERMOD can incorporate a variety of
emission source characteristics, chemical deposition properties, complex terrain, and site-specific hourly
meteorology to estimate air concentrations and deposition amounts at user-specified population
distances and at a variety of averaging times. Readers can learn more about AERMOD, equations within
the model, detailed input and output parameters, and supporting documentation by reviewing the
AERMOD Users' Guide (U.S. EPA. 2018).
Additional parameters were required to run the higher tier model, AERMOD. EPA reviewed available
literature and referenced the fenceline methodology, Draft Screening Level Approach for Assessing
Ambient Air and Water Exposures to Fenceline Communities Version 1.0 (U.S. EPA. 2022b). to select
input parameters for deposition, partitioning factors between the gaseous and particulate phases, particle
sizes, meteorological data, urban/rural designations, and physical source specifications. A full
description of the input parameters selected for AERMOD and details regarding post-processing of the
results are provided in Appendix 1.3.3.
AERMOD was run under two land categories: suburban forested and bodies of water. A limited set of
AERMOD tests suggested suburban-forest was a reasonable and appropriately health-protective default
land-cover selection when land-cover analysis is not possible. Bodies of water typically led to the
highest deposition values. Ambient air concentrations for both land categories for each OES are
presented in Appendix 1.3.3. Table 3-4 is an excerpt of the modeled annual air release data for the Use of
paints and coatings - spray application OES, 2,500 lb production volume, 95th percentile release
estimate, suburban forest land category scenario. The ambient air modeled concentrations and deposition
values are presented for two meteorology conditions (Sioux Falls, South Dakota, for central tendency
meteorology [MetCT]; and Lake Charles, Louisiana, for higher-end meteorology [MetHIGH]), 10
distances, and 3 percentiles (10th, 50th and 95th percentiles). These results indicate a maximum ambient
air concentration of 2.55 ng/m3 at 10 m from the facility and maximum deposition of 17.5 g/m2 at 30 m
from the facility for the Use of paints and coatings - spray application OES, 2,500 lb production
volume, 95th percentile release estimate, suburban forest land category scenario.
Page 65 of 638
-------
Table 3-4. Excerpt of Ambient Air Modeled Concentrations and Deposition for the Use of Paints
and Coatings - Spray Application OES, 2,500 lb Production Volume, 95th Percentile Release
Estimate, Suburban Forest Land Category Scenario
Meteorology"
Distance (m)
Concentration (ng/m3)by
Percentile
Deposition (g/m2) by Percentile
10th
50th
95th
10th
50th
95th
MetCT
10
4.98E-01
9.27E-01
1.11
3.29
7.00
8.14
MetCT
30
1.11E-01
2.84E-01
4.16E-01
2.80
5.90
7.67
MetCT
30-60
5.80E-02
1.34E-01
2.86E-01
1.22
2.67
5.78
MetCT
60
3.40E-02
9.42E-02
1.58E-01
8.46E-01
1.87
2.58
MetCT
100
1.15E-02
3.36E-02
6.45E-02
2.82E-01
6.68E-01
9.63E-01
MetCT
100-1,000
1.09E-04
5.21E-04
4.90E-03
2.21E-03
9.07E-03
8.13E-02
MetCT
1,000
5.92E-05
1.82E-04
7.95E-04
1.39E-03
3.43E-03
9.51E-03
MetCT
2,500
7.91E-06
2.39E-05
1.49E-04
1.86E-04
4.53E-04
1.78E-03
MetCT
5,000
2.29E-06
8.21E-06
4.83E-05
5.36E-05
1.71E-04
6.49E-04
MetCT
10,000
7.68E-07
2.56E-06
1.76E-05
1.85E-05
5.44E-05
2.68E-04
MetHIGH
10
5.90E-01
1.03
2.55
5.88
1.04
3.29
MetHIGH
30
1.12E-01
2.71E-01
7.05E-01
2.74
6.69
17.5
MetHIGH
30-60
4.87E-02
1.27E-01
4.32E-01
1.29
3.17
11
MetHIGH
60
2.88E-02
8.69E-02
2.23E-01
7.09E-01
2.06
5.33
MetHIGH
100
8.77E-03
3.08E-02
8.21E-02
2.13E-01
7.06E-01
1.93
MetHIGH
100-1,000
6.85E-05
4.23E-04
4.60E-03
1.61E-03
9.60E-03
1.06E-01
MetHIGH
1,000
3.25E-05
1.62E-04
6.08E-04
7.75E-04
3.68E-03
1.47E-02
MetHIGH
2,500
4.54E-06
2.52E-05
9.06E-05
1.06E-04
5.21E-04
2.19E-03
MetHIGH
5,000
1.30E-06
9.54E-06
2.87E-05
3.03E-05
1.97E-04
6.75E-04
MetHIGH
10,000
2.74E-07
4.19E-06
1.32E-05
7.09E-06
8.75E-05
2.99E-04
11 MetCT refers to meteorological conditions from Sioux Falls, South Dakota, and MetHIGH refers to meteorological
conditions from Lake Charles, Louisiana. Because the scenarios are not at real locations, they were modeled twice
using two different meteorological stations. These central tendency and high-end estimates were determined during
the development of EPA"s IIOAC.
3.3.1.2.1 Partitioning between Gaseous Phase and Particulate Phase
Dry and wet air deposition of TCEP to land and surface waters may be an important source of TCEP to
the ambient environment. Air deposition may be the result of particle deposition and/or gaseous
deposition.
There is conflicting information about the particle size of TCEP and whether TCEP is present in the gas
or particle phase. A study of offices in China suggests that the mass median aerodynamic diameters
(MMAD) of TCEP is coarse, between 4 and 5 |im, and that the contribution of TCEP is due to indoor
rather than outdoor air (Yang et al.. 2014). Another Chinese study suggests that only 22 percent of
TCEP is found among particle size fractions of dust samples less than 43 |im (He et al.. 2018c). A third
Page 66 of 638
-------
Chinese study indicates that the MMAD of TCEP is fine, between 1 and 2 |im (Cao et al.. 2019).
Schreder et al. (2016) indicates that TCEP is not detected in respirable particulate fractions (<4 |im). A
team of Canadian scientists sought to make sense of these discrepancies by examining the gas-particle
partitioning of organophosphate esters. Okeme (2018) evaluated gas-particle partitioning in indoor and
outdoor air by using a group of single-parameter and poly-parameter models. Their predictions suggest
that TCEP should be in the gas phase, contrary to measurements. Okeme (2018) suggests that the
unexpectedly high particle fractions reported in many studies is due to sampling artifact. Okeme (2018)
argues that many of the studies with high particle fractions do not account for safe sampling volumes,
and that gas-phase sorption could be contributing substantially to the mass of TCEP captured on the
filters. EPA adopted the recommendation of Okeme (2018) that many studies with the exception of
Wolschke et al. (2016) and a few others, have likely mischaracterized the gas-particle partitioning of
TCEP in air due to sampling artifact. Therefore, EPA selected a proportion of emissions in gaseous
phase of 82 percent and the proportion in particle phase of 18 percent based on Wolschke et al. (2016).
3.3.2 Water Pathway
EPA conducted systematic review to obtain concentrations of TCEP in surface water, precipitation, and
sediment. Sections 3.3.2.1, 3.3.2.3, 3.3.2.7, and 3.3.2.8 display the aggregated results of reported
monitoring and reported modeled concentrations for surface water, precipitation, and sediment found as
a result of systematic review. Section 3.3.2.4 provides surface water concentrations as a result of surface
water databases. Sections 3.3.2.5, 3.3.2.6, 3.3.2.9, and 3.3.2.10 report EPA modeled surface water and
sediment concentrations.
3.3.2.1 Geospatial Analyses of Environmental Releases
No location information is available for facilities that produce, manufacture, or use TCEP. The surface
water data from the WQP shows TCEP concentration distributed across the United States. Figure 3-3
indicates the detected water concentrations from the WQP from 1995 to 2022. Many additional sample
sites recorded non-detects, which are not shown in this figure.
Page 67 of 638
-------
Media
¦ Effluent
Finished Water
Groundwater
Hyporfieic zone
Landfill effluent
Leachate
Surface Water
Missing
Figure 3-3. Map of Nationwide Measured TCEP Water Concentrations Retrieved from the WQP
from 1995 to 2022
Source: EPA Accessible Link to Interactive Figure.
Size of the dots indicate magnitude of concentration; see Appendix 1.2.1 for more details.
3.3.2.1.1 Geospatial Analysis for Tribal Exposures
Although EPA did not identify facilities that release TCEP on or near Tribal lands, TCEP has been
detected in surface water and/or groundwater on or near Tribal lands. Groundwater samples collected in
2000 downgradient of the Norman Landfill had TCEP concentrations between 0.22 to 0.74 pg/L. Figure
3-4 indicates that the Norman Landfill was also located within a few miles from the Chickasaw Tribal
Lands in Oklahoma. The landfill closed in 1985, was covered with a clay cap, and vegetated (Barnes et
al.. 2004),
Page 68 of 638
-------
In 2018, concentrations in groundwater of up to 2.4 ug/'L. were detected at the Twenty-Nine Palms Band
of Missions Indians in Coachella, California (Figure 3-5). These concentration data were provided by
EPA's STORage and RETrieval (STORET) Data Warehouse rather than collected as part of landfill
monitoring efforts like the example above. This site was monitored again in 2019 (0.24 ug/'L) and twice
in 2021 (0.79 to 0.84 ug/L) (STORET via ( [WIS et al.. 2022)1
Page 69 of 638
-------
X X
E*)i»
Gol Cmim
\ >
J|L <«•—
V
W &
..
TEPA29
Value 2 40
Units ugA
location Surface Watei Site » 1D
Type Rivef/Stream
Ft action
Organizafcor Twenty -Ntrve
Provubet
Figure 3-5. Groundwater Concentration of TCEP Reported near Twenty-Nine Palms Reservation
near Coachella, California
Source: EPA Accessible Link to Interactive Figure.
See Appendix 1.2.1 for more details.
3.3.2.2 Measured Concentrations in Surface Water
A summary of surface water monitoring studies is provided in Figure 3-6. Six U.S. studies were
identified (five in the "US Not Specified" section and one in the "Mix Not Specified"). Sengupta et al.
(2014 reported TCEP concentrations at 581 ng/L in October 2011 and 785 ng/L in July 2011 in the Los
Angeles and San Gabriel Rivers during low flow conditions. TCEP concentrations in the Santa Clara
River, California, were recorded up to 810 ng/L during low flow events in 2013 ( laruya et al., 2016).
A Korean study found midstream concentrations of TCEP 9 times higher than upstream values (234 vs.
15.0 ng/L) (Choo et al., 2018). This study suggested that a potential cause of the elevated TCEP
concentrations was due to an industrial complex involving fiber manufacture being located near the
midstream site.
Page 70 of 638
-------
US Not Specified
Mi* N(,n 3pe
-------
LIS
4530235 - Scott et al„ 19% - US
| General Population (Background)
g Remote (Not Near Source)
NonUS
3862723-Li ct al..2017-AQ
2662833 - Mihajlovic and Fries. 2012 - DE
2662833 - Mihajlovic and Fries. 2012 - DE
im
m i
ii
2588430 - Rcgncry and Piittmann, 2010 - DE
2588430 - Regncry and Pttttmann. 2010 - DE
2598725 - Regnery and Puettmann. 2009 - DE
2598725 - Rcgncry and Pucttniann. 2009 - DE
¦
2598725 - Regncry and Pucttmann. 2009 - DE
¦
2598725 - Regnery and Puettmann. 2009 - DE
5469313 - Fries and Puttmann, 2003 - DE
m
0,01
0,1
10 100
Concentration (ng/L)
1000
Figure 3-7. Concentrations of TCEP (ng/L) in Precipitation from 1994 to 2014
3.3.2.4 Measured Concentrations in Surface Water Databases
Measured surface water concentrations were obtained from EPA's Water Quality Exchange (WQX)
using the WQP Tool, which is the nation's largest source of water quality monitoring data and includes
results from EPA's STORage and RETrieval (STORET) Data Warehouse, the U.S. Geological Survey
(USGS) National Water Information System (NWIS), and other federal, state, and Tribal sources.
The complete record of national monitoring of surface water reported by the WQP were reviewed to
summarize the prevalence of TCEP in raw surface water (NWIS et al.. 2022). Data retrieved in January
2023 included sampling dates from 2001 to 2022 and resulted in 9,892 available sample results (Figure
3-8.). Full details of the retrieval and processing of ambient surface water monitoring data from the
WQP are presented in Appendix 1.2. Figure 3-8. shows the range of TCEP concentrations detected in
surface water samples the lowest detected sample concentrations within the data set are 0.02 |ig/L. Most
of the sample records available (95%) had no level of TCEP detected above the reported detection limit
for the analysis (referred to as "non-detects"). The highest detection limit was 0.5 |ig/L. The 466
detected values ranged from 0.47 to 7.66 |ig/L, with a median of 0.23 |ig/L.
Page 72 of 638
-------
0.1 0.2 0.5 1 2
Detected TCEP Surface Water Concentration (ug/L)
Figure 3-8. Frequency of Nationwide Measured TCEP Surface Water Concentrations Retrieved
from the WQP from 2003 to 2022
The highest concentrations of TCEP detected in surface water in the United States is 7.66 |ig/L, detected
in August 2013 in Rochester, New York (NWIS via [WQP]) with a detection limit of 0.16 |ig/L. This
monitoring location is on the Genesee river at Ford Street bridge within 1,500 feet downstream of an
abandoned Vacuum Oil plant on the west bank of the Rochester's Plymouth-Exchange neighborhood.
The Vacuum Oil plant is a brownfield site that is being managed by the New York State Department of
Environmental Conservation (DEC). EPA lacks data to confirm whether Vacuum Oil is the source of
TCEP. Concentrations of up to 2.55 |ig/L have been detected in Oregon as recent as October 2020
(STORET via [WQP]). Figure 3-9 demonstrates that surface water concentrations of TCEP have been
decreasing over the past two decades.
Page 73 of 638
-------
Water Monitoring in the US by Time (excluding non-detects)
1
0.5
txQ
13
C
O
c
-------
Table 3-5 release estimates are presented based on a 2,500 lb per site-year, high-end estimate release
scenarios, the only deviation from this is the Use of paints and coatings and the Lab chemical OESs.
These deviations are due to single site throughput constraints within the models used, in these cases, the
PV of 2,500 lb/year was used to create a distribution of the possible number of sites. The 2,500 lb was
not divided by COU, rather the full 2,500 lb was considered for each COU. Because CDR reporting is
done on a per site-year basis, EPA estimated a 2,500 lb per site-year. Section 3.2 provides a summary of
the release estimates for each COU/OES. For the maximum days of release scenarios, surface water
concentrations under 7Q10 {i.e., the lowest 7-day average flow that occurs [on average] once every 10
years) flow conditions for E-FAST 2014 ranged from 5.70x 102 to 1.11 x 104 for the various exposure
scenarios. Results for VVWM-PSC are overall slightly lower for all scenarios because VVWM-PSC
accounts for additional sink effects that are not accounted for in E-FAST 2014. For more information on
E-FAST 2014 and VVWM-PSC, including information on input parameters, see Appendix 1.2.
Page 75 of 638
-------
Table 3-5. Summary of Modeled Surface Water Concentrations for the 2,500 lb, High-End Release Estimates
Life Cycle
Stage
Category
Subcategory
OES
Inputs
E-FAST 2014
WWM-PSC
Days of
Release
Estimated 7Q10
Flow (m3/day)
Daily Pollutant
Load (kg/day)
Daily Concentration
-7Q10 Oig/L)
Daily Concentration
-7Q10 Oig/L)
Manufacture
Import
Import
Repackaging
4
4,130
9.88
2,392
2,390
Processing
Incorporated
into
formulation,
mixture, or
reaction
product
Paint and
coating
manufacturing
Incorporation into
paints and
coatings - 1-part
coatings
2
3,380
35.18
10,407
10,200
Incorporation into
paints and
coatings - 2-part
coatings
1
3,380
31.89
9,436
8,280
Polymers used
in aerospace
equipment and
products
Formulation of
TCEP into 2-part
reactive resins
1
2,850
31.54
11,066
9,190
Commercial
Use
Paints and
coatings
Paints and
coatings
Use of paints and
coatings - Spray
application
2
4,130
23.26
5,631
5,590
Other use
Laboratory
chemicals
Lab chemical -
Use of laboratory
chemicals
182
4,130
0.40
96
96
7Q10 = the lowest 7-day average flow that occurs (on average) once every 10 years
Page 76 of 638
-------
3.3.2.6 EPA Modeled Surface Water Concentrations via Air Deposition (AERMOD)
A study in the lower great lakes suggested that TCEP undergoes net gas phase deposition to lakes at a
flux of-3,980 ng/m2 per day (Ma et al.. 2021). Other studies in the open ocean have suggested that the
air-water gas exchanges were dominated by volatilization from seawater to air for TCEP 146 ± 239
ng/m2 per day (Li et al.. 2017b). Moran et al. (2023) recently reported TCEP deposition to Troutman
Lake, Alaska, at magnitudes of 290 to 1300 ng/m2/day. The likely source of TCEP was suggested to be
from nearby open-air landfills and local disposal sites (Moran et al.. 2023).
EPA used IIOAC and AERMOD to estimate air deposition from facility releases and to calculate a
resulting pond water concentration near a hypothetical facility. Pond water concentrations from air
deposition were estimated for the COUs with air releases. Air deposition modeling was conducted using
IIO AC and AERMOD. Due to limitations of IIO AC in incorporating gaseous and particulate deposition,
deposition results from the AERMOD were utilized in calculating pond water concentrations. A
description of the ambient air modeling and the deposition results are provided in Section 3.3.1.2. Using
the modeled deposition rates, the TCEP concentration in pond water was calculated with the following
equations:
Equation 3-1.
AnnDep = TotDep x Ar x CF
Where:
AnnDep = Total annual deposition to water body catchment (|ig)
TotDep = Annual deposition flux to water body catchment (g/m2)
Ar = Area of water body catchment (m2)
CF = Conversion of grams to micrograms
Equation 3-2.
PondWaterConc =
AnnDep
Where:
PondWaterConc
AnnDep
Ar
Pond Depth
CF
Ar x Pond Depth
Annual-average concentration in water body (|ig/L)
Total annual deposition to water body (|ig)
Area of water body (m2); default = 10,000 m2 from EPA OPP
standard farm pond scenario
Depth of pond; default = 2 m from EPA OPP standard farm pond
scenario
Conversion of cubic meters to liters
Appendix 1.3.3 presents the range of calculated pond water concentrations for the different emission
scenarios. The highest estimated 95th percentile pond water concentration, across all exposure scenarios,
for the 2,500 lb production volume, high-end estimate was for commercial use of paints and coatings
scenario:
• 1.07xl03 |ig/L or 1,070 |ig/L at 100 m from the source; and
• 8.10 |ig/L at 1,000 m from the source.
Page 77 of 638
-------
3.3.2.7 Measured Concentrations in Wastewater
Laundry wastewater may be the primary source of TCEP to wastewater treatment plant influent and
subsequently to the aquatic environment. This theory suggests that the TCEP in the indoor environment
is transferred to indoor dust that is subsequently transferred to clothing. The dust is removed from the
clothing during laundry and this wastewater reaches the wastewater treatment plants. Not all wastewater
treatment plants are fully effective in removing TCEP, and the subsequent effluent may result in higher
concentrations in the aquatic environment (Schreder and La Guardia. 2014). Wastewater monitoring
data from multiple locations in Emeryville, California corroborates this theory, as the highest levels of
TCEP were shown to come from industrial laundry services at levels of 3.72 |ig/L in wastewater
(Jackson and Sutton. 2008). A study in Albany, New York, between 2013 and 2015 indicated mean
influent concentrations of 1,430 ng/L and effluent concentrations of 1,100 ng/L of TCEP (Kim et aL
2017). The monitoring data suggests that U.S. values of TCEP in wastewater appear to be higher than
concentrations in other high-income countries as shown in Figure 3-10.
Page 78 of 638
-------
¦¦L- Raw Influent
| Treated Effluent
m Untreated Effluent at Discharge Origin
¦ ¦ Untreated Combined Sewer Overflow
y Lognornial Distribution (CT and 90th percentile)
A Normal Distribution (CT and 90th percentile)
US
& Nan-Detect
3862000 - Kim el al.. 2017 - US
m
3862000 - Kim ct al.. 2017 - US
M
2528320 - Schroder and La Guardia. 2014 - US
Ek
2528320 - Schroder and La Guardia. 2014 - US
5469289 - Laws et al.. 2011 - US
1408465 • Jackson and Sutton. 2008 • US
1 AA
1408465 - Jackson and Sutton. 2008 - US
5743010 -1 .oraino and Pcttigrov, 2006 - US
NonUS
7002475 - Norwegian Environment, 2019 - NO
•
5428453 - Gao et al.. 2019 - SE
V
5428453 - Gao ct al.. 2019 - SE
E7 v
4457234 - Been el al.. 2017 - BE
m
5664394 - Launay et al.. 2016 - DE
¦¦CD
5664394 - Launay ct al.. 2016 - DE
4143122 - Blum ct al.. 2017-SE
^Kza
3035438 - O'Brien ct al.. 2015 - AU
mm
5469315 - Gourmclon el al.. 2010 - FR
"XI
1250860 - Rodil ct al.. 2012 - ES
mmo
1250860 - Rodil ct al.. 2012 - ES
t7V
5162720 - Meyer and Bcstcr. 2004 - DE
EA
5162720 - Meyer and Bestcr. 2004 - DE
&
8683710 - Marklund ct al. 2005 - SE
¦BEv
8683710 - Marklund el al. 2005 - SE
flv
8683710 - Marklund el al. 2005 - SE
8683710 - Marklund ct al. 2005 - SE
¦CD
8683710 - Marklund ct al. 2005 - SE
&
8683710 - Marklund el al. 2005 - SE
^37
5469313 - Fries and Puitmann. 2003 - DE
¦
5469313 - Fries and Puitmann. 2003 - DE
0.01 0.1
10 100 1000 10*4 I0A5
Concentration (ng/L)
Figure 3-10. Concentrations of TCEP (ng/L) in Wastewater from 2001 to 2018
3.3.2.8 Measured Concentrations in Sediment
Limited information was available on measured concentrations of TCEP in sediment in the United
States. Maruva et al. (2016 detected TCEP in coastal embayments at up to 6.98 ng/g dry weight in
Marina Del Ray, Los Angeles, California, in 2013. The mean sediment TCEP concentration was 2.2
ng/g with a 90th percentile value of 4.0 ng/g Maruva et al (20161. Concentrations of TCEP were
reported at a maximum of 41 ng/g in sediment samples of the Elbe River at the mouths of five tributaries
after a flooding event in Europe in August 2002 (Stachel et al.. 2005).
Page 79 of 638
-------
US Drv
4182703 - Maruya el al.. 2016 - US
¦IB General Population (Background)
Near Facility (Highly Exposed)
¦ Unknown/Not Specified
V Lognormal Distribution (CT and 90th percentile)
A Normal Distribution (CT and 90th percentile)
NonUS Drv
5305891 - Gaddha ct al., 2019 - PT
5470119 - Chokwe and Okonkwo. 2019 - ZA
V V
5469301 - Choo et al.. 2018 - KR
5740077 - Slachcl et al., 2005 - CZ,DE
2919504 - Ishikawa ct al., 1985 - IP
NonUS Wet
2935128 - Brandsma ct al,. 2015 - NL
v
0.001
0.01
0.1 1 10
Concentration (ng/g)
100
Figure 3-11. Concentrations of TCEP (ng/g) in Sediment from 1980 to 2017
Kawagoshi et al. (1999) reported TCEP concentrations up to 7,395 ng/g from waste disposal sites to the
surrounding sea in Osaka, Japan. Although this study was published in 1999, this study explains that
disposal sites may be important sources of TCEP in sediment concentrations.
3.3.2.9 EPA Modeled Sediment Concentrations (VVWM-PSC)
A summary of the benthic pore water and sediment concentrations modeled using VVWM-PSC are
summarized by COU/OES in Table 3-6. Modeled estimates are presented for the 2,500 lb production
volume, high-end estimate release scenarios. Section 3.2.2 provides a summary of the release estimates
for each COU/OES. For the maximum day of release scenarios, sediment concentrations ranged from
8.94x 102 to 5.04x 103 |ig/kg for the 2,500 lb production volume, high-end estimate release scenarios.
Page 80 of 638
-------
Table 3-6. Summary of Modeled Benthic Pore Water and Sediment Concentrations for the 2,500 lb Production Volume, High
Estimate Releases
Life Cycle
Stage
Category
Subcategory
OES
Inputs
VVWM-PSC
Days of
Release
Estimated
7Q10 Flow
(m3/day)
Daily
Pollutant
Load
(kg/day)
Benthic Pore Water
Concentration
(^g/L)
Sediment
Concentration
(ng/g)
Manufacture
Import
Import
Repackaging
4
4,130
9.88
153
894
Processing
Incorporated
into
formulation,
mixture, or
reaction
product
Paint and coating
manufacturing
Incorporation into
paints and coatings
- 1-part coatings
2
3,380
35.18
334
1,960
Incorporation into
paints and coatings
- 2-part coatings
1
3,380
31.89
152
893
Polymers used in
aerospace
equipment and
products
Formulation of
TCEP into 2-part
reactive resins
1
2,850
31.54
182
1,070
Commercial
Use
Paints and
coatings
Paints and
coatings
Use of paints and
coatings - Spray
application OES
2
4,130
23.26
177
1,040
Other use
Laboratory
chemicals
Lab chemical - Use
of laboratory
chemicals
182
4,130
0.40
90
380
Page 81 of 638
-------
For more information on the VVWM-PSC methodology, including inputs used, please see Appendix
1.2.4.
3.3.2.10 EPA Modeled Sediment Concentrations via Air Deposition (AERMOD)
EPA used AERMOD to estimate air deposition from facility releases and calculate a resulting sediment
concentration near a hypothetical facility. Sediment concentrations from air deposition were estimated
for the condition of use scenarios with air releases. Air deposition modeling was conducted using IIOAC
and AERMOD. Due to limitations of IIOAC in incorporating gaseous and particulate deposition,
deposition results from the AERMOD were utilized in calculating sediment concentrations. A
description of the modeling and the deposition results is provided above in Section 3.3.1.2. Additional
details on IIOAC and AERMOD are presented in Appendix 1.3.3. Using the modeled deposition rates,
the TCEP concentration in sediment was calculated with the following equations:
Equation 3-3.
AnnDep = TotDep x Ar x CF
Where:
AnnDep
TotDep
Ar
CF
Total annual deposition to water body catchment (|ig)
Annual deposition flux to water body catchment (g/m2)
Area of water body catchment (m2)
Conversion of grams to micrograms
Equation 3-4.
Sediment Concentration I
Where:
Sediment Cone
AnnDep
Ar
Pond Depth
Mix
Dens
AnnDep
Ar x Mix x Dens
Annual-average concentration in water body (|ig/kg)
Total annual deposition to water body (|ig)
Area of water body (m2); default = 10,000 m2 from EPA OPP
standard farm pond scenario
Depth of pond; default = 2 m from EPA OPP standard farm pond
Scenario
Mixing depth (m); default = 0.1 m
Density of sediment; default = 1,300 kg/m3 from the European
Commission Technical Guidance Document (ECB. 2003).
Appendix 1.3.31.3.3 presents the range of calculated sediment concentrations for the different emission
scenarios. Equation 3-4 is conservative as it does not include a water solubility parameter. The highest
estimated 95th percentile sediment concentration amongst all exposure scenarios was for the 2,500 lb
production volume, high-end estimate release commercial use of paints and coatings scenario:
• 1.64xl04 |ig/kg or 16,400 |ig/kg at "fenceline" population (100 m from the source); and
• 1.25xl02 |ig/kg or 125 |ig/kg at "community" population (1,000 m from the source).
3.3.3 Land Pathway
EPA conducted systematic review to obtain concentrations of TCEP in soil, biosolids, and groundwater.
Sections 3.3.3.1, 3.3.3.3, and 3.3.3.5 display the aggregated results of reported monitoring and reported
modeled concentrations for soil, sediment, and groundwater found as a result of systematic review.
Page 82 of 638
-------
Section 3.3.3.7 provides groundwater concentrations from water databases. Sections 3.3.3.2, 3.3.3.4, and
3.3.3.8 report EPA modeled and estimated soil and groundwater concentrations.
3.3.3.1 Measured Concentrations in Soil
There are no reported soil concentrations of TCEP in the United States. A research team in Germany
observed concentrations of TCEP from 5.07 to 23.48 ng/g dry weight. Snow melt appears to be a
contributor to amplified soil concentrations. The highest soil concentrations were observed 1 day after
snow melt at 23.48 ng/g, whereas soil concentrations at the same location before snowfall were below 8
ng/g. The meltwater generated at the snow surface percolated downwards due to gravity picking up
chemicals present at the snow grain edge (Mihailovic and Fries. 2012). These authors suggested that the
source of the TCEP may be due to its use in cars (Mihailovic et al.. 2011). TCEP levels ranged from
1.03 to 2.30 ng/g dry weight in Bursa, Turkey, a city known for its textile and automotive parts
manufacturing (Kurt-Karakus et al.. 2018).
3.3.3.2 EPA Modeled Soil Concentrations via Air Deposition (AERMOD)
EPA used AERMOD to estimate air deposition from facility releases and calculate a resulting soil
concentration near a hypothetical facility.
Soil concentrations from air deposition were also estimated for the COUs with air releases (see Table
3-3 for a crosswalk of COU/OES with air releases). The air deposition modeling was conducted using
IIO AC and then AERMOD. A description of the modeling and the deposition results is provided above
in Section 3.3.1.2. Using the modeled deposition rates, the TCEP concentration in soil was calculated
with the following equations:
Equation 3-5.
AnnDep = TotDep x Ar x CF
Where:
AnnDep
TotDep
Ar
CF
Total annual deposition to soil (|ig)
Annual deposition flux to soil (g/m2)
Area of soil (m2)
Conversion of grams to micrograms
Equation 3-6.
Where:
SoilConc
AnnDep
Mix
Ar
Dens
SoilConc =
AnnDep
Ar x Mix x Dens
Annual-average concentration in soil (|ig/kg)
Total annual deposition to soil (|ig)
Mixing depth (m); default = 0.1 m from the European Commission
Technical Guidance Document (TGD) (ECB. 2003)
Area of soil (m2)
Density of soil; default = 1,700 kg/m3 from TGD (ECB. 2003)
The above equations assume instantaneous mixing with no degradation or other means of chemical
reduction in soil over time and that TCEP loading in soil is only from direct air-to-surface deposition
(i.e., no runoff).
Appendix 1.3.3 presents the range of calculated soil concentrations corresponding to the emission
Page 83 of 638
-------
scenarios considered. From the table, the highest estimated 95th percentile soil concentration amongst
all exposure scenarios was for the commercial use of paints and coatings scenario:
• 1.14x 104 |ig/kg at "fenceline" population (100 m from the source); and
• 8.65x 101 |ig/kg at "community" population (1,000 m from the source)
3.3.3.3 Measured Concentrations in Biosolids
Wastewater and liquid waste treatment can result in effluent discharge to water and land application of
biosolids. A study of a wastewater treatment plant in New York reported means of combined sludge
concentrations (40.1 ng/g dry weight), ash (47.7 ng/g dry weight), and sludge cake (78.9 ng/g dry
weight) (Kim et al.. 2017). Wang et al. (2019c) reported mean and median TCEP concentrations of 10.6
ng/g dry weight and 2.46 ng/g dry weight respectively from a nationwide survey of sewage sludge
sampled in 2006-2007. TCEP was detected up to 317 ng/g dry weight in this study which collected
sewage sludge from wastewater treatment plants across 35 states in the United States (Wang et al..
2019c). Due to its persistence, and recalcitrance to anaerobic and aerobic degradation (Table 2-2.
Environmental Fate Properties of TCEPTable 2-2), it is likely that dissolved TCEP will eventually reach
surface water and groundwater via runoff after the land application of biosolids. TCEP has been found at
concentrations of 4 ng/g in Canada in biosolids (Woudneh et al.. 2015).
3.3.3.4 EPA Calculated Soil Concentrations via Biosolids
Section 2.2.3.1 indicates that TCEP will not be removed after undergoing wastewater treatment and will
be retained in effluents with a low fraction being adsorbed onto sludge.
To assess soil concentrations resulting from biosolid applications, EPA relied upon modeling work
conducted in Canada (EC/HC. 2011) that used Equation 60 from TGD (ECB. 2003). as follows:
Equation 3-7.
DC,„ _ Csludge ^ ARsludge
soil " Dsoil x BDsoil
Where:
PECsoii = Predicted environmental concentration (PEC) for soil (mg/kg)
Csiudge = Concentration in sludge (mg/kg)
ARsiudg = Application rate to sludge amended soils (kg/m2/yr); default = 0.5 from
Table A-11 of TGD
Dsoil = Depth of soil tillage (m); default = 0.2 m in agricultural soil and 0.1 m in
pastureland from Table A-l 1 of TGD
BDsoil = Bulk density of soil (kg/m3); default = 1,700 kg/m3 from Section 2.3.4 of
TGD
The concentration in sludge was assumed as 0.079 mg/kg dry weight based on Kim et al. (2017). Using
these assumptions, the estimated soil concentrations after the first year of application were 0.116 |ig/kg
in tilled agricultural soil and 0.232 |ig/kg in pastureland.
A limitation of Equation 3-7 is that it assumes no losses from transformation, degradation, volatilization,
erosion, or leaching to lower soil layers. Section 3.3.3.8 describes the potential leaching of TCEP from
landfills. Additionally, it is assumed there is no input of TCEP from atmospheric deposition and there
are no background TCEP accumulations in the soil. EPA has also assumed that there is only one
application of biosolids per year.
Page 84 of 638
-------
3.3.3.5 EPA Modeled Soil Concentrations via BST
EPA modeled soil concentrations from EPA Office of Water's (OW) Biosolids Screening Tool (BST).
The BST is a multimedia, multi-pathway, multi-receptor deterministic screening-level model that can
estimate potential exposures with land application of biosolids. The BST was peer reviewed by the EPA
Science Advisory Board in 2023 (EPA-SAB -24-001). More information is available in the BST
Supporting Documents. The BST captures a 40-year application (one application every year) of
biosolids to pastureland. Equation 3-7 calculates only one application.
Using BST, EPA estimates soil concentrations of 0.1412 mg/kg for the 2,500 lb, high-end estimate for
the Incorporation into paints and coatings - 1-part coatings OES, and 0.5293 mg/kg for the 25,000 lb,
central tendency estimate for the Incorporation into paints and coatings - 2-part reactive coatings OES.
Due to several uncertainties, EPA relied on an additional model and a few assumptions to streamline this
analysis. A key input required in the BST is the input of dry biosolids concentration pg/g in dry weight.
EPA used the RIVM SimpleTreat 4.1 model to estimate dry weight biosolids concentrations from the
environmental release estimates (see Section 3.2.1.1). For this screening analysis, only COUs with the
highest release estimates were modeled. A default annual biosolids land application rate of 1 kg/m2/year
and a TCEP biosolids concentration of 22.3 mg/kg, estimated using the SimpleTreat 4.0 wastewater
treatment plant model, were used as input to the BST.
Table 3-7. BST Mod
eled Soil
Concentrations for Incorporation into Paints and Coatings
OES
PV (lb)
Release
Estimate
(kg/day)
Release Estimate
Divided by 365
(kg/day)
SimpleTreat 4.1
Combined Sludge
(mg/kg)
BST Soil
Concentration
(mg/kg)
Incorporation into
paints and coatings
- 1-part coatings
2,500
35.2
0.20
5.95
0.1412
Incorporation into
paints and coatings
- 2-part reactive
coatings
25,000
65.9
0.75
22.3
0.5293
There are uncertainties with how long biosolids are gathered before they are eventually land applied. It
would be unreasonable to assume that the total TCEP released from a facility per year (whether it be a
few days of release or many days of release) would be applied to biosolids all at once. Thus, the release
estimate was divided by 365 to be more reflective of the land application scenario (1 application every
year).
The resulting BST soil concentrations are one order of magnitude above the soil concentrations from
modeled air deposition at 1,000 m (0.0865 mg/kg) and one order of magnitude above the soil
concentrations observed in a Germany after snow melt (0.235 mg/kg) (Mihailovic and Fries. 2012).
The BST soil concentrations are three orders of magnitude higher than the Canadian analysis described
in Equation 3-7 (0.0002 mg/kg). For a series of inputs and assumptions used in the BST analysis please
see the Supplemental File: Biosolids Screening Tool Modeling Results (U.S. EPA. 2024d).
3.3.3.6 Measured Concentrations in Groundwater
TCEP was detected in a groundwater plume downgradient (0.22-0.74 pg/L) of the Norman Landfill,
Oklahoma. The Norman Landfill is a municipal unlined landfill (subtitle D) established in 1920 and
closed in 1985 (Barnes et al.. 2004). One domestic well in Elkhart, Indiana, reported TCEP
Page 85 of 638
-------
concentrations of 0.65 to 0.74 |ig/L between 2000 and 2002. This domestic well was near Himco Dump,
a historical waste site, used for disposal until 1976 (Buszka et al.. 2009). A study from Fort Devens,
Massachusetts, reported concentrations of 0.28 to 0.81 |ig/L at monitoring wells down-gradient of a land
application facility (Hutchins et al.. 1984). These studies suggest that there is potential for TCEP to
migrate to groundwater and domestic wells from nearby non-hazardous waste landfills (e.g., Norman
Landfill) or historical waste sites (e.g., Himco Dump, Indiana, Fort Devens, Massachusetts).
us
5469289 - Laws el al.. 2011 - US
3975066 - Hopple el al.. 2009 - US
4912133 - Buszka et al.. 2009 - US
4832201 - Barnes el al.. 2008 - US
5469339 - Barnes el al.. 2004 - US
1316091 - Hutchins el al.. 1984-US
Bl Near Facility (Highly Exposed)
| General Population (Background)
V Lognormal Distribution (CT and 90th percentile)
A Normal Distribution (CT and 90th percentile)
H Non-Detect
NonUS
5428453 - Gao et al.. 2019 - SE
2579610 - Regnery et al.. 2011 - DE
2579610 - Regnery et al.. 2011 - DE
5469313 - Fries and Pulimann. 2003 - DE
5469312 - Fries and Pultmann. 2001 - DE
5469582 - Yasuhara. 1994 - JP
0.01
0,1
1 10
Concentration (ng/L)
100
1000
Figure 3-12. Concentrations of TCEP (ng/L) in the Not Specified Fraction of Groundwater from
1978 to 2017
A case study in Brazil, quantified TCEP up to 7.96 ng/L in well water samples downstream of waste
area (Cristate et al.. 2019). The waste area included bulky waste containing discarded upholstered
furniture, and mattresses. This study demonstrates that consumer discarded waste may be a source of
TCEP in landfills and subsequent migration to groundwater.
3.3.3.7 Measured Concentrations in Groundwater Databases
Data were retrieved from the WQP to characterize observed concentrations of TCEP in groundwater.
These monitored values may or may not represent locations used as a source for drinking water and are
analyzed to characterize the observed ranges of TCEP concentrations in groundwater—irrespective of
the reasons for sample collection. Data retrieved in January 2023 included sampling dates from 1995 to
2021 and resulted in 51 detected results. Figure 3-13 shows most (98%, n = 3,325) of the sample records
available had no TCEP detected above the reported detection limit for the analysis (referred to as "non-
detects"). The 51 detects had a median value of 0.21 ng/L. Full details of the retrieval and processing
groundwater monitoring data from the WQP are presented in Appendix 1.2.
Page 86 of 638
-------
c
15
c
o
jf 10-
"c
©
o
c
O
O
c
a>
cr
tt
ill
o-
0.02 0.05 0.1 0.2 0.5 I 2 5 10 20 50 100 200 500 1,004
Detected TCEP Groundwater Concentration (ug/L)
Figure 3-13. Frequency of Nationwide Measured TCEP Groundwater Concentrations Retrieved
from the WQP from 1995 to 2021
The highest concentrations of TCEP detected in groundwater in the United States is 610 |ig/L, detected
in April 2002 in Idaho. Other samples at similar locations in April 2004 were an order of magnitude
lower (2.8 to 94 |ig/L) (NWIS et al.. 2022). These estimates are from groundwater wells along the
Gooding Milner Canal in the Magic Valley. Also in 2002, TCEP was detected in groundwater in
Belleview, Florida, at a concentration of 3.5 |ig/L. A more recent value (May 2017) detected TCEP in
groundwater at a concentration of 2.4 |ig/L in New Mexico. The New Mexico monitoring location is a
well in the Four Hills Village in Albuquerque, New Mexico, which is about 1 to 2 miles from the
Kirtland Air Force Base (AFB) Landfill. Generally, based on the WQP data, concentrations of TCEP in
groundwater have been decreasing over the last two decades (Figure 3-14).
Page 87 of 638
-------
Water Monitoring in the US by Time (excluding non-detects)
3
2.5
2
-1.5
1995 2000 2005 2010 2015 2020
Time of Sampling
Figure 3-14. Time Series of Nationwide Measured TCEP Groundwater Concentrations Retrieved
from the WQP from 1995 to 2021
See Appendix 1.2.1 for more details.
3.3.3.8 EPA Modeled Groundwater Concentrations via Leaching (DRAS)
Landfills may have various levels of engineering controls to prevent groundwater contamination. These
can include industrial liners, leachate capturing systems, and routine integration of waste. However,
groundwater contamination from disposal of consumer, commercial, and industrial waste streams
continue to be a prominent issue for many landfills throughout the United States (Li et al.. 2015; Li et
al.. 2013). These contaminations may be attributed to perforations in the liners, failure of the leachate
capturing system, or improper management of the landfills. Groundwater contamination with TCEP may
occur when the chemical substance is released to landfills, underground injection wells, or surface
impoundments. Due to its physical and chemical properties (e.g., water solubility, Henry's Law
constant) and fate characteristics (e.g., biodegradability, half-life in groundwater), TCEP is anticipated
to persist in groundwater for substantially longer than in other media.
Several sources of TCEP may contribute to groundwater concentrations including industrial facility
releases and disposal of consumer products in landfills. With many manufacturing and processing uses
phased out, EPA expects environmental releases of TCEP from industrial facilities to be declining. In
fact, EPA has seen concentrations in surface water and groundwater generally declining over time.
However, environmental releases from landfills may remain (or increase). EPA considered the potential
for groundwater contamination following disposal of waste containing TCEP to landfills.
This assessment was completed using the (Hazardous Waste) Delisting Risk Assessment Software
(DRAS). DRAS was specifically designed to address the Criteria for Listing Hazardous Waste identified
Page 88 of 638
-------
in 40 CFR 261.11(a)(3), a requirement for evaluating proposed hazardous waste delisting. In this
assessment, DRAS is being utilized to determine potential groundwater concentrations of TCEP after
TCEP-containing consumer products have been disposed of into a non-hazardous waste landfill. To
understand possible exposure scenarios from these ongoing practices, EPA modeled groundwater
concentrations of TCEP leaching from landfills where TCEP or consumer products containing TCEP
have been disposed. The greatest potential for release of disposed TCEP to groundwater is from landfills
that do not have an adequate liner system.
Potential groundwater concentrations resulting from disposal of TCEP to landfills vary across landfill
loading rates and concentrations of TCEP in leachate. Estimated exposures presented here are therefore
based on varying landfill conditions. Production volumes of 2,500 lb (1,134 kg) and 2,500,000 lb.
(1,134,000 kg) are used as potential loading rates. To account for the uncertainties in estimating current
loading rates, EPA varied loading rates over four orders of magnitude. While current production
volumes are anticipated to be lower (2,500 lb), usage of TCEP was higher in the past and the leaching
occurring from landfills may be a result of past disposal practices. Furthermore, these loading rates
conservatively assume that a combination of raw TCEP and TCEP in commercial and consumer goods
all goes to a single landfill each year.
The highest leachate concentrations observed in the literature was 177 |ig/L (Masoner et al.. 2014a).
Masoner et al. (2014a) analyzed leachate concentrations from various landfills across the United States
in 2011 and 2012. In 2011, the reported range of TCEP in leachate concentrations in these landfills
ranged from 8.0><10_1 to 1.8><102 |ig/L, with a median of l.OxlO1 |ig/L and a detection frequency of 35
percent. In 2012, the maximum leachate concentration was 9.1 x 10_1 |ig/L with a detection frequency of
27 percent (Masoner et al.. 2016). A value of 3.5 |ig/L was observed from wastewater sampled from an
aerospace/aircraft modification facility in Washington State in 2021 (WSDE. 2022).
To account for the uncertainties in these estimates a range of leachate concentrations were selected for
the DRAS model from 1.0xl0~4to l.OxlO3 mg/L. The top of the range was bounded by TCEP's
solubility. DRAS calculates a weight adjusted dilution attenuation factor (DAF) based on loading rates.
The leachate concentration is divided by the DAF to calculate the groundwater concentrations. The
resulting groundwater concentrations are potential concentrations that people living within one mile of a
landfill might be exposed if the release were not identified and remediated. For more information on the
DRAS model please see Appendix 1.6.
Table 3-8. Potential Groundwater Concentrations (jig/L) of TCEP Found in Wells within
1 Mile of a Disposal Facility Determined Using the DRAS Model
Leachate
Concentration
(mg/L)
Loading Rate (kg)
1.13E-03
1.13E-04
1.13E-05
1.13E-06
1.00E-04
1.08E-06
1.01E-05
9.90E-05
9.43E-04
1.00E-03
1.08E-05
1.01E-04
9.90E-04
9.43E-03
1.00E-02
1.08E-04
1.01E-03
9.90E-03
9.43E-02
1.00E-01
1.08E-03
1.01E-02
9.90E-02
9.43E-01
1.00
1.08E-02
1.01E-01
9.90E-01
9.43
1.00E01
1.08E-01
1.01
9.90
9.43E01
1.00E02
1.08
1.01E01
9.90E01
9.43E02
Page 89 of 638
-------
Leachate
Concentration
(mg/L)
Loading Rate (kg)
1.13E-03
1.13E-04
1.13E-05
1.13E-06
1.00E03
1.08E01
1.01E02
9.90E02
9.43E03
Note: Concentrations are organized by potential loading rates (kg) and potential leachate concentrations (mg/L).
Groundwater concentrations are in ng/L.
EPA believes 2,500 lb (1.13 x 103 kg) is the most suitable production volume for current uses of TCEP.
The highest leachate concentrations observed in the literature was 177 |ig/L (Masoner et al.. 2014a).
Using these two values (1.13><103 kg and 0.177 mg/L), the expected groundwater concentrations are
1.08xl0~3 to 1.08><10~2 Hg/L. Disposals of TCEP containing materials were likely a few orders of
magnitude higher than current levels when the TCEP production volume was higher and TCEP was
more commonly used. If taking a production volume of 250,000 lb (1.13x105 kg), and 0.177 mg/L
leachate concentration the expected groundwater concentrations are 9.90x 10~2 to 9,90x10 1 |ig/L.
These estimates are within the range of the groundwater concentrations reported in the monitoring
literature: 0.28 to 0.81 |ig/L in a study from Fort Devens, MA monitoring wells down-gradient of a land
application facility (Hutchins et al.. 1984). 0.65 to 0.74 |ig/L between 2000 and 2002 at a domestic well
in Elkhart, Indiana near Himco Dump (Buszka et al.. 2009). In addition, these values are below
groundwater concentrations of 2.4 |ig/L detected in May 2017 in New Mexico reported in the WQP. The
New Mexico monitoring location is a well in the Four Hills Village in Albuquerque, New Mexico,
which is about 1 to 2 miles from the Kirtland AFB Landfill.
3.4 Concentrations of TCEP in the Indoor Environment
TCEP - Concentrations in the Indoor Environment (Section 3.4):
Key Points
EPA evaluated the reasonably available information for concentrations of TCEP in the indoor
environment. The key points are summarized below:
• The indoor environment exposure characterization focused on consumer uses, disposals, and
background exposures of TCEP.
o Indoor air monitoring data show TCEP in particulate or vapor/gas form with
concentrations primarily between 1 x 10~2 and 1x 104 ng/m3.
o Indoor dust is an important exposure pathway for TCEP. EPA found monitoring data
showing a range of TCEP concentrations in indoor dust in residential spaces, public
spaces, and vehicles, with concentrations as high as 167,532 ng/g in homes.
The indoor environment exposure characterization focuses on consumer uses, disposals, and background
exposures of TCEP. In addition to the contribution from consumer uses, indoor environment TCEP
concentrations were estimated from ambient contributions for air.
Note that indoor air and dust concentrations from consumer uses are presented in Section 5.1.2.
For more information on TCEP indoor monitoring and reported indoor modeling data, please see:
• Environmental Monitoring Concentrations Reported by Media Type (U.S. EPA. 2024i);
Page 90 of 638
-------
• Environmental Monitoring and Biomonitoring Concentrations Summary Table (U.S. EPA.
2024h);
• Data Quality Evaluation Information for General Population, Consumer, and Environmental
Exposure. (U.S. EPA. 2024x); and
• Data Extraction Information for General Population, Consumer, and Environmental Exposure
(U.S. EPA. 2024r).
3.4.1 Indoor Air Pathway
3.4.1.1 Measured Concentrations in Indoor Air
The indoor air monitoring data indicates indoor air concentrations primarily between 1 x 1CT2 and 1 x 104
ng/m3 ranges. One study indicated particulate concentrations of TCEP of up to l.lxlO7 ng/m3 max in
PM2.5 (Wallner et al.. 2012). This study may have had issues with sampling artifacts due to the use of
glass filters as described by Okeme (2018) (see Section 3.3.1.2 for more details). There was only one
study on vapor/gas in the United States. Dodson et al. (2017) reported a 95th percentile concentration of
37 ng/m3 TCEP in vapor/gas.
Page 91 of 638
-------
US Combined Vapor/Gas and Particulate
5432871 - Dodson ct al- 2019 - US
NonUS Combined Vapor/Qas and Particulate
4659643 - Okeme ct al.. 2018 - CA
4285929 - He et al.. 2018 - AU
4285929 - He ct al.. 2018 - AU
4659643 - Okcmc ct al., 2018 - CA
5165777 - Lazarov ct al, 2015 - BE
3005686 - Takeuchi ct al, 2015 - JP
2560628 - Makinen ct al., 2009 - Fl
2560628 - Miikinen ct al.. 2009 - Fl
697390 - Kanazawa ct al., 2010 - JP
632484 - Ohura ct al., 2006 - JP
4731349 - Ingcrowski ct al.. 2001 - DE
3012534 - La Guardia and Hale, 2015 - US
3012534 - La Guardia and Hale, 2015 - US
2539068 - Bradman et al., 2014 - US
4292129 • Deng ct al., 2018 - CN
5163827 - Wong ct al., 2018 - SE
1927779 - Saito et al.. 2007 - JP
1927779 - Saito et al..2007-JP
5755270 - Dodson el al, 2017 - US
US Particulate
NonUS Particulate
US Vapor/Gas
NonUS Vapor/Gas
4292133 - Persson ct al.. 2018 - SE
5083520 - Sha ct al., 2018 - SE
5083520 - Sha ct al., 2018 - SE
3357642 - Xu el al., 2016 -NO
3604490 - Tokumura et al.. 2017 - JP
IOA-4
*
¦ Public Space
¦ Residential
¦i Vehicle
V Lognormal Distribution (CT and 90th percentile)
A Normal Distribution (CT and 90th percentile)
$ Non-Detect
^7
I V V
mmsmm
¦7 V
W
/A
i V V
I V V
I V V
0,1 I 10
Concentration (ng/m3) (pt 1)
100
1000
10*4
Page 92 of 638
-------
(continued)
NonUS Vapor/Gas
2537005 - Fromme ct al.. 2014 - DE
788335 - Bcrgh et al., 2011 - SE
788335 - Bcrgh ct ah. 2011 - SE
5469670 - Liiotlgo and Ocsimun, 2016 - SE
1249459 - Bergh et al.. 2011 - SE
779503 - Haitmann ct al.. 2004 - CH
779503 - Hartmann ct al.. 2004 ¦ CH
1949033 - Yoshida ct al.. 2006 - JP
789515 - Olake et al.. 2004 - JP
1598712 - Olake ct al.. 2001 - JP
IOA-4
g Public Space
| Residential
Vehicle
V Lognonnal Distribution (CT and 90th percentile)
I V V
0.1 I 10
Concentration (ng/m3) (pt 2)
10*4
Figure 3-15. Concentrations of TCEP (ng/m3) in Indoor Air from 2000 to 2016
3.4.1.2 Measured Concentrations in Personal Air
Two studies measured TCEP in personal air in the U.S. Personal air refers to the area within the
breathing zone. Schreder et al. (2016) conducted a study on white-collar workers in urban, suburban,
and rural areas of Washington State. Participants were instructed to wear an Institute of Occupational
Medicine (IOM) sampler affixed to a shirt collar within the breathing zone continually during a 24-hour
day during normal activities, including at home and at work, traveling to and from home and work,
shopping, and socializing, and to wear or hang the sampler at breathing zone level during sleep.
Schreder et al. (2016) reported mean and maximum inhalable (>4 |im) TCEP concentrations of 19.1
ng/m3 and 77.8 ng/m3 respectively, detected in 8/9 participants. La Guardia and Hale (2015) conducted a
study measuring flame retardants among the personal air of four gymnastics coaches at their workplace
and their homes. TCEP was not detected in the personal air of these coaches. Okeme et al. (2018)
reported a median personal air concentration of three Canadian office workers of 34 ng/m3.
Polydimethylsiloxane (silicone rubber) brooches were used for the sampling methodology, and the three
participants wore the samplers for 7 days.
US Particulate
NonUS Particulate
NoriUS Vapor/Gas
General Population (Background)
3222316 - Schrcdcr ct al., 2016 - US
5017615 - Okeme ct al.. 2018 - CA
3357642 - Xu ct al., 2016 - NO
0.001
III
1
0.1 I
Concentration (ng/m3)
Figure 3-16. Concentrations of TCEP (ng/m3) in Personal Inhalation in General Population
(Background) Locations from 2013 to 2016
3.4.1.3 EPA Modeled Indoor Concentrations as a Ratio of Ambient Air
IIOAC calculates a mean and high-end indoor air concentration based on the outdoor/ambient air
concentration and the mean and high-end indoor-outdoor ratios. In IIOAC, indoor-outdoor ratios of 0.65
and 1 are used for the mean and high-end ratios, respectively. The indoor-outdoor ratio of 0.65 is used to
calculate indoor air concentrations corresponding to the mean outdoor air concentration for each
Page 93 of 638
-------
potentially exposed population. The indoor-outdoor ratio of 1 is used to calculate the indoor air
concentration corresponding to the 95th percentile of outdoor air concentration of each potentially
exposed population.
IIOAC was used as a tier 1 screening model before estimating ambient exposures via AERMOD.
Results of IIO AC are presented in Appendix 1.3.
3.4.1.4 Reported Modeled Concentrations in Indoor Air
Shin et al. (2014) reported TCEP emission rates in a whole house of 48.417 mg/day. Emission rate refers
to the amount of chemical emitted per unit time. Shin et al. (2014) utilized fugacity-based indoor mass
balance models to estimate whole-house emission rates of various SVOCS including TCEP.
Page 94 of 638
-------
3.4.2 Indoor Dust Pathway
3.4.2.1 Measured Concentrations in Indoor Dust
us
Mix
NonUS
4161719 - Hoffman cl al., 2017 - US
5163584 - Phillips cl al., 2018 - US
6968217 - Shin cl al„ 2019 - US
3012534 - La Guardia and Hale. 2015 - US
3012534 - La Guardia and Hale. 2015 - US
2343712 - Siaplcton cl al.. 2014 - US
2528320 - Schredcr and La Guardia, 2014 - US
2539068 - Bradman cl al., 2014 - US
2215665 -Shin el al.. 2014-US
1676728 - Fang el al.. 2013 - US
1676728 - Fang el al.. 2013 - US
3864462 - Castorina cl al„ 2017 - US
5184432 - Tan el al.. 2019 - CN.US
5043338 - Veldzquez-Gdmez et al . 2019 - ES
5043338 - Velazquez-G6mcz et al.. 2019 - ES
5043338 - Veld/qucz-G6niez et al.. 2019 - ES
5163693 - Raniakokko et al.. 2019 - FI
5165944 - Liu and Mabury, 2019 - CA
5412073 - Giovanoulis et al.. 2019 - SE
3223090 - Langer el al.. 2016 - DK
3223090 - Langer el al., 2016 - DK
4292121 - Christia etal .2018 - GR
4292129 - Deng et al.. 2018 - CN
4292133 - Persson et al.. 2018 - SE
3862555 - Zhou et al.. 2017 - DE
3862555 - Zhou et al.. 2017 - DE
3862555 - Zhou et al.. 2017 - DE
4285929 - He et al.. 2018 - AU
0.01
||H Residential
¦ Public Space
¦¦¦ Vehicle
V Lognormal Distribution (CT and 90th percentile)
A Normal Distribution (CT and 90lh percentile)
no
—mr.
I v v
I
I V V
I" g o
V
' V
32E
I7V
W
IV V
IV V
I V V
I vv
mmzmrnm
~ v
IV
0.1
J7V
I V V
10 100 1000 10*4
Concentration (ng/g) (pi 1)
I0A5
I0A6
10A7
Page 95 of 638
-------
(continued)
NonUS
NonUS
4285929 - He ct al.,20l8-AU
4285929 -He ct al..2018-AU
4292136 - Larson ct al.. 2018 - SE
3005686 - Takcuchi ct al.. 2015 - JP
3357642 - Xu cl al., 2016 - NO
4178500 - Kim and Tanabe. 2017 - KR
4178500 - Kim and Tanabc. 2017 - KR
4433160 - Kadcmoglou ct al., 2017 - GB.NO
1313395 - Wallncr ct al., 2012 - AT
3604490 - Tokumura ct al.. 2017 - JP
3975074 - Sugeng ct al,. 2017 - NL
4433160 - Kadcmoglou ct al. 2017 - GB
4829235 - Ail Bamai ct al.. 2018 - JP
1927602 -Ali et al..20l2-NZ
2537005 ¦ Fromme ct al.. 2014 - DE
2540527 - Brandsma ct al.. 2014 - NL
2540527 - Brandsma ct al . 2014 - NL
3350460 - Coelho et al.. 2016 - PT
5164389 - Brommcr ct al.. 2012 - DE
788335.Bcrghctal .20II - SE
788335 - Bergh et aL. 2011 - SE
1927614 - Van den Eede ct al.. 2012 - BE.ES.RO
2542290 • Tajima ct al., 2014 - JP
2543095 - Fan ct al.. 2014 - CA
3015040 - Mizouehi ct al.. 2015 - JP
5469392 - Bastiacnscn ct al.. 2019 - JP
5469670 - Luongo and Ocsiman. 2016 - SE
697390 - Kanazawa ct al . 2010 - JP
2919501 - Marklund ct al.. 2003 - SE
2919501 - Marklund et al.. 2003 - SE
0.01
| Rcsideiilial
¦ Public Space
Vehicle
-g»i Other
V Lognoruial Distribution (CT and 90th percentile)
A Normal Disiribution (CT and 90lh percentile)
MS?
^J7
I V V
I V V
w
~ V
17 V
10 100 1000 10*4
Concentration (ng/g) (pi 2)
I0A5
10A7
I Residential
IVehicle
A Normal Disiribution (CT and 90th percentile)
2919501 - Marklund et al.. 2003 - SE
2919501 - Marklund et al.. 2003 - SE
4731349 - Ingerowski et al.. 2001 - DE
0.01
1
HZ
0.1
10 100 1000 10*4
Concentration (ng/g) (pi 3)
10*5
Figure 3-17. Concentrations of TCEP (ng/g) in Indoor Dust from 2000 to 2019
Concentrations of TCEP in dust were significantly higher in facilities with napping equipment (e.g.,
foam beds and mats) made from foam (Bradman et al.. 2014). Correlations between organophosphate
Page 96 of 638
-------
esters in dust and consumer products containing foams, furniture, and electronics strongly implicate
household items as sources of these chemicals (Abafe and Martincigh. 2019). In the United States,
concentrations of TCEP in dust are reported at 50.2 ng/g in houses and up to 1,080 ng/g in cars (Fang et
al.. 2013). Phillips et al. (2018) reported maximum concentrations of TCEP of 167,532 ng/g and a
geometric mean of 864.1 ng/g in North Carolina homes from September 2014 to April 2016 as part of
the Toddler's Exposure to SVOCs in the Indoor Environment (TESIE) study. A study of the Center for
the Health Assessment of Mothers and Children of Salinas (CHAMACOS) cohort in California reported
similar concentrations of TCEP as the TESIE cohort. It found that TCEP levels in dust are significantly
associated with the presence of extremely worn carpets (Castorina et al.. 2017).
3.4.2.2 Reported Modeled Concentrations in Indoor Dust
Castorina et al. (2017) reported modeled oral doses of 0.064 |ig/kg-day for pregnant women via
residential indoor dust in Salinas Valley, California. Schreder et al. (2016) reported 50th percentile
modeled intakes for children (82.8 ng/day) and adults (41.4 ng/day). Ingerowski et al. (2001). a low-
quality study, reported a range of dust intakes of from 0.2 to 2 |ig/day.
Rantakokko et al. (2019) modeled inhalation, dermal, and oral intakes of TCEP in children from indoor
dust. 50th percentile intakes were highest for dust ingestion (2.9 ng/kg-day) vs. dermal absorption (1.3
ng/kg/day) and inhalation (0.023 ng/kg-day). This suggests that for children's exposure to dust, oral
routes may be the most important avenue of exposure. Kademoglou et al. (2017) modeled adult and
toddler daily dust intakes from European homes and offices. They reported mean toddler dust intakes of
14.195 ng/kg/day for the high intake rate and 3.549 ng/kg/day in houses located in the United Kingdom.
Adult intakes were higher in houses (0.624 ng/kg bw with high intake rate) vs. offices (0.0214 ng/kg bw
with high intake for 8 hours spent in offices). The highest observed modeled dust intakes (1.38 |ig/kg-
day) were reported for children at a kindergarten in Hong Kong (Deng et al.. 2018b).
Page 97 of 638
-------
4 ENVIRONMENTAL RISK ASSESSMENT
EPA assessed environmental risks of TCEP exposure to aquatic and terrestrial species. Section 4.1
describes the environmental exposures through surface water, sediment, soil, air, and diet via trophic
transfer. Environmental hazards for aquatic and terrestrial species are described in Section 4.2, while
environmental risk is described in Section 4.3. Updates since the draft risk evaluation within the
Environmental Risk Assessment section include: (1) Integrated updated peer-reviewed literature search
in February 2024 to fill data gaps for environmental hazard; (2) Revised concentrations) of concern
(COCs) for chronic aquatic hazards for both sediment and surface water compartments; (3) Calculated
risk estimates from soils to include soil concentrations from the BST analysis.
Page 98 of 638
-------
4.1 Environmental Exposures
Environmental Exposures (Section 4.1):
Key Points
EPA evaluated the reasonably available information for environmental exposures of TCEP to aquatic
and terrestrial species. The key points of the environmental exposure assessment are summarized
below:
• EPA expects the main environmental exposure pathways for TCEP to be surface water,
sediment, and soil. The ambient air exposure pathway was also assessed for its contribution
via deposition to these media.
• TCEP exposure to aquatic species through surface water and sediment were modeled to
estimate concentrations near industrial and commercial uses. These results were compared to
measured concentrations of TCEP from databases (i.e., WQP) or published literature from a
variety of locations.
o Modeled data estimate surface water concentrations in the low thousands of ppb (Table
4-11) and pore water concentrations low hundreds of ppb (Table 4-12) near industrial and
commercial uses.
o Monitoring data show TCEP surface water concentrations in the United States generally
decreasing over the last two decades.
o While EPA does not expect TCEP to bioaccumulate in higher trophic levels in the food
web, biomonitoring from the published literature show TCEP in the tissue of several
aquatic species including fish in the Great Lakes and harbor seals in San Francisco Bay.
o EPA also estimated fish tissue concentrations by COU using the modeled water releases
from industrial and commercial uses.
• TCEP exposure to terrestrial species through soil, air, and surface water was also assessed
using modeling and monitoring data.
o TCEP exposure to terrestrial organisms occurs primarily through diet via the soil
pathway, with deposition from air to soil being a source. Exposure through diet was
assessed through a trophic transfer analysis, which estimated the transfer of TCEP from
soil through the terrestrial food web using representative species.
o TCEP exposure to terrestrial organisms from surface water ingestion is typically
ephemeral. Therefore, the trophic transfer analysis for terrestrial organisms assumed
TCEP exposure concentrations for wildlife water intake are equal to TCEP soil
concentrations for each corresponding exposure scenario.
o Direct exposure of TCEP to terrestrial receptors via air was not assessed quantitatively
because dietary exposure was determined to be the driver of exposure to wildlife. The
contribution of TCEP exposure from inhalation relative to the ingestion exposure route is
not expected to drive risk because of dilution associated environmental conditions.
4.1.1 Approach and Methodology
Soil and surface water are the major environmental compartments for TCEP (see Section 2.2.2). The
environmental exposure assessment focuses on TCEP concentrations in surface water, sediment, and soil
Page 99 of 638
-------
as these are the media used to determine risks to aquatic and terrestrial organisms (see Section 4.3).
Ambient air was also assessed for its contribution via deposition to these media.
Monitoring information for aquatic and terrestrial species are presented in Sections 4.1.2 and 4.1.3
below. Reported monitoring information on environmental media (e.g., surface water, sediment, air) are
presented in Section 3.3. When available, measured TCEP concentrations from databases (i.e., WQP) or
published literature were used as comparative exposure concentrations for risk quotient (RQ)
calculations and are presented in Section 4.3.
EPA utilized various models to assess the environmental concentrations resulting from the industrial and
commercial release estimates (see Section 3.3). These models are E-FAST 2014, VVWM-PSC, IIOAC,
and AERMOD. Additional information on these models is available in Section 3.3. TCEP surface water
concentrations (ppb) were modeled by E-FAST 2014 and VVWM-PSC. TCEP pore water and benthic
concentrations were modeled using VVWM-PSC as described in Section 3.3.2.9. TCEP concentrations
in soil, surface water, and sediment via air deposition at the community level (1,000 m from the source)
were modeled as described in Sections 3.3.3.2, 3.3.2.6, and 3.3.2.10, respectively. Reported and
modeled surface water and sediment concentrations were used to assess TCEP exposures to aquatic
species.
Measured and modeled soil concentrations were utilized to assess risk to terrestrial species via trophic
transfer (see Section 4.1.4). Specifically, trophic transfer of TCEP and potential risk to terrestrial
animals was based on modeled soil data from AERMOD and concentrations reported within Mihailovic
and Fries (2012). Potential risk to aquatic dependent wildlife utilized surface water concentrations
modeled via VVWM-PSC for each COU in combination TCEP fish concentrations calculated using the
whole body BCF reported within (Arukwe et al.. 2018). Exposure factors for terrestrial organisms used
within the trophic transfer analyses are presented in Section 4.1.4. Application of exposure factors and
hazard values for organisms at different trophic levels is detailed within Section 4.3 and utilized
equations as described in the U.S. EPA Guidance for Developing Ecological Soil Screening Levels (U.S.
EPA. 2005a).
For more information on TCEP monitoring data in aquatic and terrestrial species, please see the
following supplemental documents:
• Environmental Monitoring Concentrations Reported by Media Type (U.S. EPA. 2024i);
• Environmental Monitoring and Biomonitoring Concentrations Summary Table (U.S. EPA.
2024h);
• Data Quality Evaluation Information for General Population, Consumer, and Environmental
Exposure. (U.S. EPA. 2024x); and
• Data Extraction Information for General Population, Consumer, and Environmental Exposure
(U.S. EPA 2024rV
4.1.2 Exposures to Aquatic Species
4.1.2.1 Measured Concentrations in Aquatic Species
A graphical survey of TCEP concentrations in fishes within reasonably available published literature
(seven studies) is presented in Figure 4-1. Guo et al. (2017b) measured concentrations of TCEP in fish
samples in the Great Lakes Basin using the Great Lakes Fish Monitoring and Surveillance Program
(GLFMSP) sampling protocol. TCEP was found in more than 50 percent of the fish samples at a
geometric mean of 13.3 ng/g lipid, including lake trout (Salvelinus namaycush) or walleye (Sander
Page 100 of 638
-------
vitreus). The lipid-based concentrations of TCEP in Lake Erie fish were significantly higher than those
of the other four Great Lakes. These concentrations are in line with lipid-based concentrations from
Sundkvist et al. (2010). who measured TCEP in mussels (Mytilus edulis), herring (Clupeidae), eelpout
(Zoarces vivipcirus), salmon (Scilmo salar), and perch (Perca fluviatilis) in Swedish lakes and coastal
areas.
TCEP has been recorded in the blubber of harbor seal (Phoca vitalina) within the San Francisco Bay at a
median concentration of 3.4 ng/g (Sutton et al.. 2019). Sutton et al. (2019) indicated that blubber might
not be a good indicator of exposure to hydrophilic phosphate-based flame retardants due to degradation
and metabolism. Two European studies present lipid concentrations of TCEP in aquatic mammals at
similar levels to the lipid concentrations in fish shown above (Sala et al.. 2019; Hallanger et al.. 2015).
Mix Lipid
—| General Population (Background)
I Remote ( Not Near Source)
Near Facility (Highly Exposed)
A Normal Distribution (CT and 90th percentile)
^ Lognormal Distribution (CT and 90th percentile)
3985267 - Guo el al.. 2017 - CA.US - Other
A A
NonUS Lipid
5164.108 - Sanlin et al.. 2016 - ES - Whole Organism
5162922 - Hallanger cl al., 2015 - NO ¦ Other
2586188 - Sundkvist cl al.. 2010 - SE - Muscie/Filci
2586188 - Sundkvist el al.. 2010 - SE - Muscle/Filet
2586188 - Sundkvist et al., 2010 - SE - Muscle/Filet
&
w
NonUS Wet
5469301 - Choo ct al., 2018 - KR - Liver
5469301 - Choo et al., 2018 - KR - Muscle/Filet
5469301 - Choo et al.. 2018 - KR - Other
5469297 - McGoldrick ct al., 2014 - CA - Other
2935128 - Brandsma ct al.. 2015 - NL - Other
6992056 - Evenset et al., 2009 - NO - Liver
6992056 - Evenset et al., 2009 - NO - Muscle/Fillet
6992056 - Evenset ct al., 2009 - NO ¦ Whole Organism
0.001
W
fl7
S77
Dv
0.01
0.1
100
1000
Concentration (ng/g)
Figure 4-1. Measured Concentrations of TCEP (ng/g) in Aquatic Species - Fish from 2003 to 2016
4.1.2.2 Calculated Concentrations in Aquatic Species
In addition to considering monitoring data from published literature, EPA modeled concentrations in
fish for each industrial and commercial release scenario (Table 4-1). Concentrations of TCEP in fish
were calculated by multiplying the VVWM-PSC modeled surface water concentrations for each
industrial and commercial releases scenario by the BCF of 0.34 L/kg (Arukwe et al.. 2018) (Table 2-2).
These conservative whole fish TCEP concentrations were utilized within the screening-level assessment
for trophic transfer as described in Section 4.1.4.
Page 101 of 638
-------
Table 4-1. TCEP Fish Concentrations Calculated from VVWM-PSC Modeled Industrial and
Commercial TCEP Releases
Production
Volume (lb/year)
Release
Distribution"
Surface Water
Fish
Scenario Name
Concentration
(^g/L)
Concentration
(ng/g)
Import and repackaging
2,500
High-End
2,350
799
Incorporation into paints
and coatings - 1-part
coatings
2,500
High-End
10,000
3,400
Incorporation into paints
and coatings - 2-part
reactive coatings
2,500
High-End
8,150
2,771
Use of paints and coatings -
spray application
2,500
High-End
5,500
1,870
Formulation of TCEP
2,500
High-End
9,040
3,073
containing reactive resin
Laboratory chemicals
2,500
High-End
96
32
11 Production volume of 2,500 lb TCEP/year uses high-end estimates (95th percentile for all COUs except the laboratory
chemicals COU that uses the 1st percentile).
These calculated whole fish results are one to three orders of magnitude higher than the reported fish
concentrations in Guo et al. (2017b). who reported a geometric mean of 35.6 ng/g lipid in Lake Erie.
Guo et al. (2017b) also reported a geometric mean concentration of TCEP in Great Lakes water of
4.64x 10~4 |ig/L via Venier et al. (2014). while Arukwe et al. (2018) used a water concentration of
7.75><102 |ig/L to derive the BCF within laboratory-controlled experiments. The current TCEP surface
water concentrations modeled via VVWM-PSC are one to two orders of magnitude greater that values
reported in Arukwe et al. (2018); however, it is important to consider that modeled concentration are
intended to represent COU-based source release concentrations.
4.1.2.3 Modeled Concentrations in the Aquatic Environment
E-FAST 2014 was used to estimate total TCEP surface water concentration within lotic (i.e., flowing)
systems and represents TCEP concentration within the water column. The days of exceedance modeled
in E-FAST 2014 are not necessarily consecutive and could occur throughout a year at different times.
Days of exceedance is calculated as the probability of exceedance multiplied by the total modeled days
of release. While both E-FAST 2014 and VVWM-PSC consider dilution and variability in flow, the
VVWM-PSC model can estimate a time-varying surface water concentration, partitioning to suspended
and settled sediment, and degradation within compartments of the water column. VVWM-PSC considers
model inputs of physical and chemical properties of TCEP (i.e., Kow, Koc, water column half-life,
photolysis half-life, hydrolysis half-life, and benthic half-life), allowing EPA to model predicted pore
water and sediment concentrations.
The VVVM-PSC model utilized relatively low stream orders (i.e., depth of 2 m) as a conservative
approach for modeling stream reach. Results within PSC are reported as the maximum concentration
value of the investigated chemical over the specified averaging periods (e.g., 1-day, 3-day, etc.) as well
as a time-series graph of surface water and benthic pore water concentrations (U.S. EPA. 2019e). TCEP
surface water concentrations (ppb) were modeled by E-FAST 2014 and VVWM-PSC and are presented
in Table 4-11 for each COU at a production volume of 2,500 lb per year. TCEP pore water concentration
modeled by VVWM-PSC are presented within Table 4-11 and Table 4-12, respectively.
Page 102 of 638
-------
EPA used 110AC and AERMOD to estimate air deposition from facility releases and calculate a
resulting pond water concentration near a hypothetical facility. Pond water concentrations from air
deposition were estimated for the COUs with air releases (Table 4-9). AERMOD results indicate air
deposition to water are not drivers of risk and have significantly reduced TCEP concentrations when
compared to TCEP when modeled within the water column, pore water, and sediment modeling via E-
FAST 2014 and VVWM-PSC. For example, the highest estimated 95th percentile pond water
concentration from annual deposition from air to water, across all exposure scenarios, was 8.1 |ig/L for
the Commercial use of paints and coatings scenario at an annual production volume of 2,500 lb. This
highest modeled concentration (8.1 |ig/L) within a pond at 1,000 m from a point source was
approximately 12 times lower than the lowest surface water concentration modeled using VVWM-PSC
(96 |ig/L as a maximum 1-day average concentration for the laboratory chemicals scenario at an annual
production volume of 2,500 lb). Although the IIOAC and AERMOD were applied to a generic farm pond
setting to calculate concentrations of TCEP in pond surface water and pond sediment, these models do not
account for media exchange of the chemical of interest as VVWM-PSC does.
4.1.3 Exposures to Terrestrial Species
4.1.3.1 Measured Concentrations in Terrestrial Species
Two studies (see Figure 4-2) have reported concentrations of TCEP and a TCEP metabolite bis(2-
chloroethyl) phosphate (BCEP) in bird eggs (Guo et al.. 2018; Stubbings et al.. 2018). From these two
studies the mean concentration of TCEP in birds by wet weight is 5.3 ng/g with a 90th percentile value
of 9.7 ng/g. BCEP was among the most abundant metabolites (0.38 to 26 ng/g ww) in bald eagle
(Haliaeetus leacocephalus) eggs. These values are results of the Michigan Bald Eagle Biosentinel
Program archive that sampled bald eagles in the Great Lakes Region from 2000 to 2012.
US Wet
NonUS Wet
B I General Population (Background)
| Remote (Not Near Source)
¦Ml Near Facility (Highly Exposed)
a Non-Detect
V Lognomial Distribution (CT and 90th percentile)
5166846 - Guo el al.. 2018 - US - Bloodg
5166846 - Guoet al.. 2018 - US - Egg (whole)
w
2823276 ¦ Huberet al., 2015 ¦ NO - Egg (whole)
4181327 - Chen ct al.. 2012 - CA - Egg (whole)
4931691 - Greaves and Letcher. 2014 - CA - Blood
4931691 - Greaves and Letcher, 2014 - CA - Liver
4931691 - Greaves and Letcher, 2014 - CA - Other
4931691 - Greaves and Letcher, 2014 - CA - Adipose Tissue
4931691 - Greaves and Letcher. 2014 - CA - Egg (yolk)
4931691 - Greaves and Leichcr, 2014 - CA - Muscle/Filet
07
w
M
9
NonUS Dry
5017003 - Monclus et al., 2018 - ES - Feathers
2542346 - Eulaers el al., 2014 - NO - Feathers
2935128 - Brandsma cl al., 2015 - NL - Egg (whole)
IOA-6
V V
IOA-4
0.01 1
Concentration (ng/g)
100
10*4
Figure 4-2. Measured Concentrations of TCEP (ng/g) in Terrestrial Species - Bird from 2000 to
2016
Page 103 of 638
-------
Aston et al. (1996) reported TCEP in pine needles (Pinus ponderosa) at six out of nine collection sites in
the Sierra Nevada Foothills in the mid-1990s with a geometric mean TCEP concentration of 142 ng/g
and a range of 10 ng/g to 1,950 ng/g (Figure 4-3). Although the source of the TCEP is unknown, the
authors suspected that concentrations may have been due to aerial transport and deposition from nearby
point sources such as incinerators. Samples reported within Aston et al. (1996) were collected in 1993
and 1994 with concentrations from this study representing a period with significantly higher
concentrations of TCEP in production and use (see Section 1.1.1).
us Wet
5469881 - Aston et al.. 1996 - US - Foliage
| Remote (Not Near Source)
V Lognormal Distribution (CT and 90th percentile)
0.01
0.1
1 10 100 1000
Concentration (ng/g)
10*4
Figure 4-3. Measured Concentrations of TCEP (ng/g) in the Wet Fraction of Terrestrial Species -
Plant in Remote (Not Near Source) Locations from 1993 to 1994
4.1.3.2 Modeled Concentration in the Terrestrial Environment
The contribution of exposure risk from inhalation relative to the ingestion exposure route is not expected
to drive risk because of dilution associated environmental conditions (U.S. EPA. 2003a. b). In addition,
TCEP is not persistent in air due to its short half-life in the atmosphere (ti/2 = 5.8 hours) and because
particle-bound TCEP is primarily removed from the atmosphere by wet or dry deposition (U.S. EPA.
2017a). Air deposition to soil modeling is described in Section 3.3.3.2. EPA determined the primary
exposure pathway for terrestrial organisms is through soil via dietary uptake via trophic transfer. As
described in Section 3.3.3.2, IIOAC and subsequently AERMOD were used to assess the estimated
release of TCEP via air deposition from specific exposure scenarios to soil. Estimated concentrations of
TCEP that could be in soil via air deposition at the community level (1,000 m from the source) exposure
scenarios have been calculated and are presented in Appendix H.2.
4.1.4 Trophic Transfer Exposure
Trophic transfer is the process by which chemical contaminants can be taken up by organisms through
dietary and media exposures and transferred from one trophic level to another. EPA has assessed the
available studies collected in accordance with the 2021 Draft Systematic Review Protocol (U.S. EPA.
2021a) relating to the biomonitoring of TCEP.
TCEP is released to the environment by various exposure pathways (see Figure 2-1). The exposure
pathway for terrestrial organisms is through soil; deposition of TCEP from air to soil is the primary
exposure pathway. A secondary source of TCEP contamination in soil is from the application of
biosolids. However, the concentration of TCEP in soil from biosolids is two orders of magnitude less
than the TCEP soil concentration from air deposition (see Section 3.3). Therefore, biosolid application is
not expected to drive risk within the terrestrial environment. The exposure pathway for water includes
runoff from soil (e.g., after a rain event), deposition from air, and direct releases from water treatment
plants. Sediment TCEP concentrations determined by VVMW-PSC modeling range from 2.6- to 108.8-
fold greater than surface water concentration across all COUs (see Section 3.3.2.9). Indicating that
sediment acts as a sink for TCEP and a source of elevated exposure to TCEP through the dietary
exposure pathway for higher trophic levels in the water column that feed on benthic organisms. Trophic
magnification is not expected in the water column or terrestrial environments but may occur where
TCEP concentrations are high (i.e., in the benthic zone) (Table 2-2).
Representative avian and mammal species are chosen to connect the TCEP transport exposure pathway
via terrestrial trophic transfer from earthworm (Eisenia fetida) uptake of TCEP from contaminated soil
Page 104 of 638
-------
through invertivore avian (American woodcock [Scolopax minor]) and mammal (short-tailed shrew
[Blarina brevicauda]) species, to the American kestrel (Falco sparverius) that feeds on invertebrates,
avian, and small terrestrial vertebrates.
American woodcocks primarily feed on invertebrates with a preference for earthworms. When
earthworms are not available, other soil invertebrates and a small proportion of vegetation may be
consumed. Depending on the location and season, earthworms may comprise 58 to 99 percent of
American woodcock diet (U.S. EPA 1993b). Short-tailed shrews primarily feed on invertebrates with
earthworms comprising approximately 31 percent (stomach volume) to 42 percent (frequency of
occurrence) of their diet. American kestrels have a varied diet that includes invertebrates and vertebrates
(mammal, avian, and reptile). The proportion of prey type will vary by habitat and prey availability. For
trophic transfer analysis, the American kestrel diet comprised equal proportions of the three
representative prey species (i.e., one-third earthworm, one-third American woodcock, and one-third
short-tailed shrew), which approximates the dietary composition of the American kestrel winter diet
reported in Meyer and Balgooven (1987). The calculations for assessing TCEP exposure from soil
uptake by earthworms and the transfer of TCEP through diet to higher trophic levels are presented in
Section 4.3.1.10. Because surface water sources for wildlife water ingestion are typically ephemeral, the
trophic transfer analysis for terrestrial organisms assumed TCEP exposure concentration for wildlife
water intake are equal to soil concentrations for each corresponding exposure scenario (U.S. EPA
2003a. b).
The representative semi-aquatic terrestrial species is the American mink (Mustela vison), whose diet is
highly variable depending on their habitat. In a riparian habitat, American mink derive 74 to 92 percent
of their diet from aquatic organisms, which includes fish, crustaceans, birds, mammals, and vegetation
(Alexander. 1977). Similar to soil concentrations used for terrestrial organisms, the highest modeled
surface water TCEP concentrations with a production volume of 25,000 lb/year was used as a surrogate
for the TCEP concentration found in the American mink's diet in the form of both water intake and a
diet of fish. For trophic transfer, fish concentrations shown in Table 4-1 are used in conjunction with
trophic transfer calculations in Section 4.3.1.1.
Page 105 of 638
-------
ED
Figure Legend
Negligible
» Low/Slow
Moderate
High/Fast/Strong
Very HigWRapid/Strong
Partitioning/T ransportation
~ T ransform ation/Degradation
Wastewater Facility
Figure 4-4. Trophic Transfer of TCEP in Aquatic and Terrestrial Ecosystems"
11 The diagram demonstrates uptake from media to biota and trophic transfer through the food web (blue arrows).
The width of the arrows shows relative chemical transport between biota or media. Within the aquatic
environment, the benthic zone is bounded by dashed black lines from the bottom of the water column to sediment
surface and subsurface layers. The depth that the benthic environment extends into subsurface sediment is site
specific. The conceptual model illustrates BCFs, BSAFs, and TMFs for aquatic organisms as shown in Appendix
F.2.6. Food intake rates (FIRs) are shown for terrestrial vertebrates.
4.1.5 Weight of Scientific Evidence Conclusions for Environmental Exposures
4.1.5.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the
Environmental Exposure Assessment
Concentrations of TCEP in environmental and biological media are expected to vaiy. Release from
industrial facilities, indoor sources, and long-range transport may all contribute to concentrations of
Page 106 of 638
-------
TCEP in the environment. Determining the source apportionment of TCEP from each is complex.
Proximity to facilities and other sources is likely to lead to elevated concentrations compared to
locations that are more remote. No manufacturing or processing facility locations were identified for
releases to TCEP. The inability to locate releases in proximity to facility locations contributes to a layer
of uncertainty when selecting model input parameters that are typically informed by location (e.g.,
meteorological data, land cover parameters for air modeling, flow data for water modeling).
Limited monitoring data are available for aquatic and terrestrial species in the United States. In addition,
monitoring data collected in previous years when production volume and associated releases of TCEP
into the environment are expected to have been higher than they are currently and expected to be in the
future. When considering older monitoring data and monitoring data from international sources, there
are uncertainties associated with using these data because it is unknown whether those sampling sites are
representative of current sites within the United States. Recent and future estimated levels of TCEP in
the area may be lower than past levels due to reported reductions in releases over time. The predicted
concentrations may be lower than concentrations that consider more years of releases or releases
associated with higher production volumes.
In modeling environmental concentrations of TCEP, EPA acknowledges the conservative nature of the
E-FAST 2014 model and the additional refinement provided by the VVWM-PSC model. Water dilution
models can be used to determine the concentration of a chemical in the surface water after a source
releases the chemical into a water body. Because the E-FAST 2014 model default values encompass
either a combination of upper percentile and mean exposure parametric values, or all upper percentile
parametric values, the resulting model predictions represent high-end exposures estimates. A simple
dilution model such as E-FAST 2014 provides exposure estimates that are derived from a simple mass
balance approach and does not account for partitioning between compartments within a surface water
body or degradation over time in different media, parameters which are relevant to TCEP. For these
reasons, EPA utilized a two-tier approach by complementing the E-FAST 2014 modeling with more
refined estimates from the PSC model to describe further environmental exposures.
When modeling using E-FAST 2014, EPA assumed that primary treatment removal at POTWs occurred
with 0 percent removal efficiency. EPA recognizes that this is a conservative assumption that results in
no removal of TCEP prior to release to surface water. Section 2.2.1 and Appendix F.2.5.2 discusses the
recalcitrance of TCEP to wastewater treatment systems. This assumption reflects both the uncertainty of
the type of wastewater treatment that may be in use at a direct discharging facility and the TCEP
removal efficiency in that treatment.
EPA used a combination of chemical-specific parameters and generic default parameters when
estimating surface water, sediment, soil, and fish-tissue concentrations. For estimated soil concentrations
from air deposition, specifically, EPA recognizes that different default parameters for gaseous vs.
particle partitioning, may result in concentrations of a higher magnitude. However, EPA used central
tendency, high production volume, and high-end, central tendency production volume values to
characterize the variability within and across scenarios. To estimate soil concentrations, EPA also used
central tendency and high-end meteorological inputs.
Comparison of model outputs with monitored values offers one way to ground truth the combination of
model inputs and outputs used. EPA compared monitoring and modeled surface water, sediment, soil,
and fish-tissue concentration estimates. Estimates of fish-tissue concentrations are further discussed in
Section 5.1.3.4.2. In summary, EPA compared monitored and modeled fish tissue concentrations and
found modeled fish concentrations were two to three orders of magnitude higher than those reported for
Page 107 of 638
-------
whole fish within published literature (see Section 4.1.2.2). The conservative approach for calculated
fish tissue concentrations presented in Section 4.1.2.2 was utilized for trophic transfer analysis to semi-
aquatic mammals (see Section 4.3.1.10). In comparison to measured values reported within published
literature, these calculated values should be viewed as organisms with direct proximity to source of
TCEP release as calculated using VVWM-PSC.
EPA conducted modeling of TCEP concentrations in surface water, pore water, and sediment based on
the assumption that releases entered lotic (flowing) aquatic systems. Although EPA did not consider the
potential impact of persistence and longer-term sinks in lake and estuary environments, localized
deposition of TCEP within 1,000 m from hypothetical release sites from air to soil, water, and sediment
were modeled for each applicable COU via IIOAC and AERMOD.
Page 108 of 638
-------
4.2 Environmental Hazards
Environmental Hazards (Section 4.2):
Key Points
EPA evaluated the reasonably available information for environmental hazard endpoints associated
with TCEP exposure. The key points of the environmental hazard assessment are summarized below:
• Aquatic species hazard:
o Aquatic hazard data were available for TCEP for six species of fish, four invertebrate
species, and five algae species.
o To estimate hazards (mortality) from acute exposures, EPA supplemented the empirical
data with hazard predictions from an EPA predictive tool, Web-based Interspecies
Correlation Estimation. These data were used with the empirical fish and daphnid data to
create a species sensitivity distribution (SSD) and calculate a TCEP concentration of
concern (COC) for acute exposures of aquatic species (16,700 ppb) representing the
lower 95th percentile of an HC05 (Table 4-6).
o EPA applied Web-based Interspecies Correlation Estimation (Web-ICE) to empirical
hazard data on algae to create an SSD and calculated a COC for aquatic plants of 66,000
ppb (Table 4-6).
o EPA also calculated a COC for chronic exposures (survival of yellow catfish) to aquatic
species (2.8 ppb) using empirical fish data (Table 4-6).
• Terrestrial species hazard:
O Terrestrial hazard data for TCEP were available for soil invertebrates, mammals, and
avian species.
O Based on empirical toxicity data for nematodes and earthworms, the chronic hazard
threshold for terrestrial invertebrate is 612 mg/kg soil (Table 4-7).
O Empirical toxicity data for mice and rats were used to estimate a chronic toxicity
reference value (TRY) for terrestrial mammals of 44 mg/kg-bw/day (Table 4-7).
4.2.1 Approach and Methodology
During the scoping process, EPA reviewed potential environmental hazards associated with TCEP
exposure and identified 14 sources of environmental hazard data shown in Figure 2-10 of Final Scope of
the Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) CASRN115-96-8 (U.S. EPA. 2020b).
EPA completed the review of environmental hazard data/information sources during risk evaluation
using the data quality evaluation metrics and the data quality criteria described in the 2021 Draft
Systematic Review Protocol (U.S. EPA 2021a). Studies were assigned an overall quality determination
of high, medium, low, or uninformative.
EPA assigned an overall quality determination of high or medium to 23 acceptable aquatic toxicity and
17 acceptable terrestrial toxicity studies. For the aquatic studies on vertebrates and invertebrates, seven
species had appropriate endpoint concentrations (LC50) for assessing the effects from acute exposures
of TCEP. Five empirical hazard values were available for aquatic plants, represented by green algae and
diatom species. The modeling approach, Web-ICE, Version 3.3, can both predict toxicity values for
environmental species that are absent from a dataset and provide a more robust dataset to estimate
Page 109 of 638
-------
toxicity thresholds. EPA used Web-ICE to supplement empirical acute hazard data for aquatic
organisms. Details outlining the method are included in Appendix G. For terrestrial species, all mammal
studies were from mice and rats used as human health model organisms. These studies were used to
calculate a toxicity reference value (TRV) for mammals, which is expressed as doses in units of mg/kg-
bw/day. Although the TRV for TCEP is derived from laboratory mice and rat studies, because body
weight is normalized, the TRV can be used with ecologically relevant wildlife species to evaluate
chronic dietary exposure to TCEP. Representative wildlife species chronic hazard thresholds are
evaluated in the trophic transfer assessments using the TRV.
4.2.2 Aquatic Species Hazard
Toxicity to Aquatic Organisms
EPA assigned an overall quality determination of high or medium to 23 acceptable aquatic toxicity
studies. These studies contained relevant aquatic toxicity data for: brine shrimp (Artemia sp.), diatoms
(Phaeodactylam tricornutum) and {Skeletonema costatam), green alga (Dunaliella salina), (Platymonas
subcordiformis), and (Raphidocelis sabcapitata), Japanese seabass (Lateolabrax macalatas), Japanese
medaka (Oryzicis latipes), Manila clam (Ruditcipes philippinarum), mrigal carp (Cirrhimts mrigala), mysid
shrimp (Neomysis awatschensis), rainbow trout (Oncorhynchus mykiss), water flea (Daphnia magna),
yellow catfish (Pelteobagrus fidvidraco), and zebrafish (Danio rerio). EPA identified 12 aquatic toxicity
studies, displayed in Table 4-2, Table 4-3, and Table 4-4 as the most relevant for quantitative
assessment. The remaining studies were represented by results at a sub-organ or mechanistic level,
considered to be separated from direct population level effects or did not demonstrate effect(s) at the test
concentrations employed within their study concentrations gradients. The Web-ICE application was
used to predict LC50 toxicity values for 46 aquatic species (22 fish, 1 amphibian, 9 aquatic
invertebrates, and 14 benthic invertebrates) using empirical acute toxicity data from rainbow trout,
zebrafish, and Daphnia magna (Raimondo and Barron. 2010). Invertebrate and vertebrate species with
empirical hazard values (n = 7) and predicted species (n = 46) toxicity data were subsequently used to
calculate the distribution of species sensitivity to acute TCEP exposure (Appendix G). The Web-ICE
application was used to predict EC50 values for three algal species using empirical hazard values from
one freshwater green algae species {Raphidocelis sabcapitata) and two marine diatoms (.Phaeodactylam
tricornutum and Skeletonema costatam). The five empirical and three predicted species hazard values
were subsequently used to calculate the distribution of species sensitivity for aquatic plants exposed to
TCEP (Appendix G).
Aquatic Vertebrates
Relevant acute toxicity studies for fish that included LC50 data were assigned an overall quality
determination of high for three 96-hour static condition fish toxicity studies (Zhang et al.. 2024;
Alzualde et al.. 2018; Life Sciences Research Ltd. 1990a) evaluating the median LC50 from exposure to
TCEP. The acute 96-hour LC50 values for fish were 249 mg/L for rainbow trout (Life Sciences
Research Ltd. 1990a). 279 mg/L for zebrafish embryo (Alzualde et al.. 2018). and 46 mg/L for juvenile
Japanese seabass (Zhang et al.. 2024). The LC50 study for rainbow trout did not meet the assumptions
of the Probit test. Therefore, a non-linear interpolation was used to approximate the LC50 value. The
zebrafish embryo study by Alzualde et al. (2018) used a non-linear regression test (sigmoidal dose-
response curve) to calculate the LC50. Zhang et al. (2024) used log probability curves estimate 96-hourr
LC50 values and 95 percent confidence intervals (CIs) for Japanese seabass.
The ChV is the geometric mean of the lowest-observed-effect-concentration (LOEC) and no-observed-
effect-concentration (NOEC). The overall quality determination for relevant studies with acute exposure
duration ChV values were high for two 96-hour studies for rainbow trout and zebrafish (Alzualde et al..
2018; Life Sciences Research Ltd. 1990a). The 96-hour rainbow trout had a ChV of 70.7 mg/L for
Page 110 of 638
-------
mortality (Life Sciences Research Ltd. 1990a) while the 96-hour zebrafish embryo had a ChV of 139.7
mg/L for development and growth (Alzualde et al.. 2018).
Six studies on fishes were available and represented TCEP exposure with subchronic and chronic
durations ranging from 14 to 120 days. All six studies were assigned an overall quality determination of
high for the apical assessment endpoints of regulatory interest (i.e., impaired growth, survival, or
reproduction). The shortest exposure duration was a study on Japanese Medaka encompassing 14-day
TCEP exposures across approximately 9 days of embryo development followed by approximately 5
days of larval development (Sun et al.. 2016). Daily TCEP renewal of 90 percent of the treatment and
solvent control waters was conducted every 24 hours throughout the exposure period with no analytical
verification of TCEP concentrations. The duration of this experimental exposure covering all of
embryogenesis and 5 days of larval development representing sensitive lifestages for fishes. Reporting
nominal concentrations, the 14-day exposure to TCEP resulted in a ChV of 0.559 mg/L for development
and growth with significant differences in length compared to control groups (Sun et al.. 2016). Mean
hatchability, reported as a percent, was not significantly different among TCEP treatments and
decreased with increased TCEP treatment concentrations of 1.25 mg/L and 6.25 mg/L resulting in
hatchability of 92.6 ± 7.4 and 90.6 ± 5.8, respectively, compared to the control percent hatchability of
96.3 ± 3.7. Similarly, increasing TCEP treatment concentrations of 1.25 mg/L and 6.25 mg/L resulting
in non-significant increases in percent gross abnormality rates of 5.8 ± 2.4 and 5.9 ± 2.4, respectively,
compared to the control gross abnormality rate of 3.2 ± 1.0. Authors reported TCEP-exposed larvae
were observed to have increasing trends in swimming speeds for both dark and light phases, however,
these quantified movements were not significantly different from control.
Juvenile mrigal carp were exposed to TCEP for 21 days at nominal concentrations of 0.04, 0.2, and 1
mg/L with daily TCEP renewal of 75 percent of the treatment and control waters (Sutha et al.. 2020).
Structural abnormalities based on histology were observed within the lowest TCEP treatment
concentration (0.04 mg/L) for gill, liver, and kidney tissues with the severity of abnormalities increasing
with increasing TCEP treatment concentrations. Authors described the greatest incidences of
abnormalities occurring within gill and liver tissues. For example, gill tissue abnormalities included but
were not limited to: epithelial lifting, hyperplasia, and degeneration of cells in primary lamellae; while
abnormalities observed within liver tissue included but were not limited to: necrosis, pyknotic nuclei,
and increased sinusoids vessels. Sutha et al. (2020) also observed significant differences in plasma
thyroid hormones and antioxidant enzyme activities at the lowest treatment concentration (0.04 mg/L).
Although authors reported non-quantified behavioral changes associated with TCEP exposure compared
to control treatments the study did not record any changes in growth or survival from a 21-day TCEP
exposure.
Two studies were conducted with 30-day TCEP exposures to juvenile yellow catfish with daily 66
percent water replacement and analytically verified TCEP concentrations (Hu et al.. 2022: Zhao et al..
2021). Water samples were collected twice a week for analytical verification of TCEP using Liquid
chromatography-mass spectrometry. Hu et al. (2022) reported nominal TCEP treatment concentrations
of 0, 0.001, 0.01, and 0.1 mg/L with analytically verified concentrations of 0, 0.00087 ± 0.00012,
0.00924 ± 0.00131, and 0.08915 ± 0.00463 mg/L, respectively. Hazard values are represented by
analytically verified TCEP concentrations as reported by the authors and rounded to represent 0.0009,
0.009, and 0.09 mg/L. Survival was significantly decreased at a TCEP treatment concentration of 0.09
mg/L compared to the solvent (0.01% DMSO) control with a NOEC of 0.009 mg/L resulting in a ChV
of 0.028 mg/L. Final weight and length were significantly reduced compared to control groups at a
TCEP concentration of 0.009 mg/L. Final length and weight at the lowest TCEP treatment concentration
of 0.0009 mg/L were not significantly different from the control treatment. Specific growth rate was
Page 111 of 638
-------
significantly higher at 0.0009 mg/L and was significantly reduced compared to control at both the 0.009
mg/L and 0.09 mg/L treatment concentrations. Gill histology indicated shortened secondary gill lamellae
and epithelial hyperplasia at the lowest nominal TCEP concentration of 0.0009 mg/L, however, authors
did not perform severity scoring or report the proportions of individuals affected Zhao et al. (2021).
Severity scoring and the precent of individuals affected was documented with histological analysis of
the liver conducted by Hu et al. (2022). Liver tissues displayed no histological changes for the control
and 0.0009 mg/L TCEP treatment group, while the 0.009 mg/L treatment group displayed cytoplasmic
vacuolization (10-50% affected), cellular peripheral nucleus (<10% affected), and karyolysis (<10%
affected). Authors reported observing increase incidences of karyolysis (10-50% affected), cellular
degeneration (10-50% affected), and focal necrosis (<10% affected) within the liver at the 0.09 mg/L
TCEP treatment.
Two studies were conducted on zebrafish with 120-day TCEP exposures beginning at the embryo stage
and measuring outcomes within sexually mature adults (Hu et al.. 2023) and outcomes from resulting
embryos and larvae (Wang et al.. 2022). The solvent DMSO was used at a final concentration of 0.0001
percent within both solvent control and TCEP treatment groups. TCEP and solvent control treatments
had daily 50 percent water replacement and water samples were collected before and after renewal at
day 119 of the exposure period for analytical verification using liquid chromatography-mass
spectrometry. Authors reported nominal TCEP treatment concentrations of 0, 0.0008, 0.004, 0.020,
0.100 mg/L and analytically verified concentrations after renewal of 0, 0.00072 ± 0.00014, 0.00372 ±
0.00031, 0.01868 ± 0.00155, and 0.09377 ± 0.00629 mg/L, respectively. Hazard values are represented
by analytically verified TCEP concentrations as reported by the authors and rounded to represent
0.0007, 0.004, 0.019, and 0.09 mg/L. The concentration of TCEP within newly fertilized embryos
resulting from parental exposures was reported for each concentration within Wang et al. (2022). Body
mass and length for adults exposed to TCEP for 120 days were significantly reduced compared to
controls at TCEP concentrations of 0.004 mg/L and greater (Hu et al.. 2023). Adult hepatosomatic
index, percent weight of the liver relative to body weight, was significantly reduced in all treatment
concentrations compared to control treatment (Hu et al.. 2023). The effects of adult 120-day TCEP
exposure on newly fertilized embryos and larvae reared in TCEP free water were reported within Wang
et al. (2022). with apical endpoints recorded at 120 hours post fertilization (hpf) related to survival,
hatch rate, heart rate, body length, and incidence of malformations. Heart rate was significantly
increased compared to control groups at an adult 120-day TCEP exposure concentration of 0.09 mg/L.
Survival at 120 hpf was significantly different from control for adult 120-day TCEP exposure
concentrations at and above 0.004 mg/L. Categories of malformations observed within 120 hpf larvae
included pericardial edema, yolk sac edema, tail deformation, and spinal curvature. The incidence of
malformations at 120 hpf in larvae were significantly greater than control for all adult 120-day TCEP
exposure concentrations.
Page 112 of 638
-------
Table 4-2. Aquatic Vertebrate Environmental Hazard Si
tudies for TCE
>
Duration
Test Organism (Species)
Endpoint(s)
Hazard Values
(mg/L)
Geometric
Mean6 (mg/L)
Effect
Citation
(Data Evaluation Rating)
Acute
Fish: rainbow trout
(Oncorhvnchus mvkiss)
96-hour LC50
96-hour NOEC/LOEC
249
50/100
70.7
Mortality
(Life Sciences Research Ltd.
1990a) (Hieh)
Fish: Japanese Seabass
(Lateolabrax maculatus)
96-hour LC50
46
—
Mortality
(Zhane et al.. 2024) (Hieh)
Fish: zebrafish embryo
(Danio rerio)
96-hour LC50
279
-
Mortality
(Alzualde et al.. 2018) (Hieh)
96-hour EC50
96-hour NOEC/LOEC
118
114/171
139.7
Developmental/
Growth
Subchronic
/Chronic
Fish: Japanese medaka
('Oryzias latipes)
14-day NOEC/LOEC
0.25/1.25
0.559
Developmental/
Growth
(Sun et al.. 2016) (Hieh)
Fish: mrigal carp
('Cirrhinus mrigal a)
21-day LOEC
0.04
-
Renal/Kidney
(Sutha et al.. 2020) (Heh)
Hepatic/Liver
Respiratory
Fish: yellow catfish
(Pelteobagrus fulvidraco)
30-day NOEC/LOEC
0.009/0.09
0.028
Mortality
(Hu et al.. 2022) (Hieh)c'#
0.0009/0.009
0.0028
Developmental/
Growth
0.0009/0.009
0.0028
Hepatic/Liver
Fish: zebrafish (Danio
rerio)
120-day NOEC/LOEC
120-day LOEC
0.0007/0.004
0.0016
Developmental/
Growth
(Hu et al.. 2023) (Hieh)'#
0.0007
-
Hepatic/Liver
Fish: zebrafish embryo/
larvae (Danio rerio)
120-day F0 exposure; F1
NOEC/LOEC
LOEC
LOEC
0.0007/0.004®
0.0016e
Mortality
(Wane et al.. 2022) (Heh)'#
0.09e
-
Cardiovascular
0.0007'#
-
Reproductive/
Teratogenic
" Hazard value represented by the corresponding "Endpoint" in the adjacent column (e.g., LC50, NOEC, LOEC).
b Geometric mean of definitive values only.
c Hu et al. (2022) has the same data and data quality ranks for mortality and erowth/development outcomes reported within Zhao et al. (2021).
J Hazard values are represented by analytically verified concentrations as reported by the authors.
e TCEP concentrations represent adult 120-day exposures with progeny (embryo/larvae) reared for 120 hours in water containing no TCEP. Maternal transfer of TCEP
was confirmed from analytical chemistry performed on eggs from each treatment and control groups.
Page 113 of 638
-------
Amphibians
No amphibian studies were reasonably available to assess potential hazards from TCEP exposure.
However, modeled data from Web-ICE predicted a bullfrog (.Lithobates catesbeicmus) 96-hour LC50
with a geometric mean of 264 mg/L from two surrogate species. Therefore, amphibians are accounted
for within the Web-ICE and species sensitivity distribution (SSD) results.
Aquatic Invertebrates
Three studies were available on aquatic invertebrates and represented with acute exposures to different
species and chronic exposures to Daphnia magna (Table 4-3). Two related studies conducted on
Daphnia magna represent both acute and chronic exposures of TCEP. The 48-hour TCEP exposure
resulted in an EC50 for immobilization of 171 mg/L with a LOEC for immobilization at 48-hours of 117
mg/L (Torav Research Center. 1997a). A chronic study with 21-day exposure of TCEP to Daphnia
magna resulted in a NOEC and LOEC for the cumulative number of offspring produced per parent of 10
mg/L and 17 mg/L, respectively (Torav Research Center. 1997c). Working with marine species, Zhang
et al. (2024) used log probability curves to estimate 48-hr LC50 values for brine shrimp and 96-hr LC50
values for mysid shrimp and Manila clams.
Table 4-3. Aquatic Invertebrate Environmental Hazard Studies for TCEP
Duration
Test Organism
(Species)
Endpoint
Hazard Values
(mg/L)"
Geometric
Mean6 (mg/L)
Effect
Citation
(Data Evaluation
Rating)
Acute
Aquatic
Invertebrate:
brine shrimp
(Artemia sp.)
48-hour LC50
97
Mortality
(Zhang et al.. 2024)
(High)
Aquatic
Invertebrate:
mysid shrimp
(Neomysis
awatschensis)
96-hour LC50
40
Mortality
Mollusk: Manila
clam (Ruditapes
philippinartim)
96-hour LC50
312
Mortality
Aquatic
Invertebrate:
Daphnia magna
48-hour EC50
171
-
Immobilization
(Torav Research
Center. 1997a)
(High)
4 8-hour
NOEC/LOEC
76/117
94.2
Immobilization
Chronic
Aquatic
Invertebrate:
Daphnia magna
21-day LC50
83.1
-
Mortality
(Torav Research
Center. 1997c)
(High)
21-day EC50
29.6
-
Reproduction
21-day
NOEC/LOEC
10/17
13
Reproduction
" Hazard value represented by the corresponding "Endpoint" in the adjacent column (e.g., LC50, NOEC, LOEC)
h Geometric mean of definitive values only.
Aquatic Plants
Two studies were available and represent TCEP exposures to two species of marine diatoms, two
species of marine green algae, and one species of freshwater algae (Table 4-4). One study represented a
72-hour exposure of TCEP to the freshwater green algae, Raphidocelis sabcapitata (Torav Research
Center. 1997b). Exposure of TCEP to green algae for 72 hours resulted in a NOEC and LOEC for the
growth inhibition of 65 mg/L and 116 mg/L, respectively (Torav Research Center. 1997b). Zhang et al.
Page 114 of 638
-------
(2024) conducted TCEP exposures on marine microalgae and used log probability curves to estimate 96-
hr EC50 values for growth on two diatom species, Phaeodcictylam tricornutum and Skeletonema
costatam, and two marine green algae species, Dunaliella salina and Platymonas subcordiformis.
Table 4-4 Aquatic Plant Environmental Hazard Studies for TC
EP
Test Organism
(Species)
Endpoint
Hazard
Values
(mg/LK
Geometric
Mean6 (mg/L)
Effect
Citation (Data
Evaluation Rating)
Diatom:
Phaeodactyhim
tricornutum
96-hour EC50
76
Growth
(Zhana et al.. 2024)
Diatom: Skeletonema
costatam
96-hour EC50
353
-
Growth
Green Algae:
Dunaliella salina
96-hour EC50
265
-
Growth
(High)
Green Algae:
Platymonas
subcordiformis
96-hour EC50
395
Growth
Green Algae:
Raphidocelis
subcapitata0
72-hour EC50
212
-
Growth
(Torav Research
Center. 1997b) (Hiah)
72-hour
NOEC/LOEC
65/116
87
Growth
a Hazard value represented by the corresponding ""Endpoint" in the adjacent column (e.g., LC50, NOEC, LOEC)
h Geometric mean of definitive values only.
cTest species formerly known as Selenastrum capricornutum and Pseudokirchneriella subcapitata.
^4.2.3 Terrestrial Species Hazard
EPA assigned an overall quality determination of high or medium to 17 acceptable terrestrial toxicity
studies. These studies contained relevant terrestrial toxicity data for two Norway rat (Rattus norvegicus)
strains (F334 and Sprague-Dawley), two mouse (Mus musculus) strains (CD-I IGS and B6C3F1), one
earthworm (Eiseniafetida), and one nematode (round worm; Caenorhabditis elegans). EPA identified a
total of seven terrestrial toxicity studies, displayed in Table 4-5, as the most relevant for quantitative
assessment.
Terrestrial Vertebrates
Five relevant chronic toxicity studies for terrestrial vertebrates that included no-observed-effect level
(NOEL) and/or lowest-observed-effect level (LOEL) data were assigned an overall quality
determination of high or medium with reproduction, mortality, and/or neurotoxicity (e.g., lesions to
hippocampus) endpoints for rodents (n = 4) and thyroid effects for the single avian toxicity study. One
study with a medium overall quality determination was for the reproduction endpoints reported within
Matthews et al. (1990). Mortality endpoints within the same study received an overall quality
determination of high.
Similarities among mammalian studies with ecologically relevant, population-level effects were
observed. Of the three studies that included mice, two studies resulted in LOEL values. Reproductive
effects (NOEL =175 mg/kg, LOEL = 700 mg/kg) due to reduced sperm count was shown in Matthews
et al. (1990). An initial dose gradient for a single dose reproduction study found that the lowest test dose
with mortality effects in mice was LOEL = 1,000 mg/kg (Hazleton Laboratories. 1983). Additionally,
ataxia and tremors were noted shortly after dosing of the mice, which may be related to neurotoxicity.
Page 115 of 638
-------
Male rats were more sensitive (NOEL = 88 mg/kg, LOEL =175 mg/kg) to TCEP exposure through the
oral route for mortality endpoints than females (NOEL =175 mg/kg, LOEL = 350 mg/kg) (Matthews et
al.. 1990). The 2-year studies for neurotoxicity (degenerative lesions of cerebrum and brain stem) and
mortality endpoints showed a NOEL of 44 mg/kg and a LOEL of 88 mg/kg (NTP. 1991b). A 60-day
Sprague-Dawley rat study also resulted in neurotoxicity with lesions in the hippocampus (Yang et al..
2018a). These studies indicate that neurotoxicity of the brain may be a mode of action (MO A) for TCEP
exposures in rodents.
For avian species, one high-quality study was available for the American kestrel (Fernie et al.. 2015).
The study reported statistically significant increases in the plasma free thyroid hormones
triiodothyronine (T3) and thyroxine (T4) (LOEL = 0.0025 mg/kg-bw/day) with no effects on body
weight or food consumption from 21-day TCEP exposure through the diet.
Soil Invertebrates
Relevant chronic toxicity studies for soil invertebrates included two studies that were assigned an overall
quality determination of high. The earthworm had a NOEL of 0.1 mg/kg soil and a LOEL of 1.0 mg/kg
soil at 3, 7, and 14 days of exposure to TCEP that showed a significant dose response relationship with
degradation of the digestive tract and exfoliation of the typhlosole (Yang et al.. 2018b). The nematode
study results show a NOEL of 500 mg/kg soil and a LOEL of 750 mg/kg soil at 3 days exposure to
TCEP for reduced growth and shortened lifespan, and an LC50 of 1,381 mg/kg soil at 6 days exposure
to TCEP CXu et al.. 2017).
Terrestrial Plants
No terrestrial plants studies were available to assess potential hazards from TCEP exposure.
Page 116 of 638
-------
Table 4-5. Terrestrial Organisms
Environmental Hazard Studies Used
For TCEP
Duration
Test Organism
Endpoint
Hazard Values
(mg/kg rb
Geometric Meanc
(mg/kg)
Effect
Citation
(Data Evaluation Rating)
Mammals
Chronic
F344/N rats
(Rattus norvegicus)
2-year NOEL/LOEL
44/88
62.2
Neurotoxicity/
mortality
(NTP. 1991b) (Hieh)
16-week
NOEL/LOEL
Female: 175/350
Male: 88/175
247.5
124.1
Mortality
(Matthews et al.. 1990)
(High)
B6C3F1 mice (Mas
musculus)
16-week NOEL/
LOEL
175/700
495.0
Reproduction
(Matthews et al.. 1990)
(Medium)
Sprague-Dawley rat
(Rattus norvegicus)
60-day NOEL/LOEL
50/100
70.7
Neurotoxicity
(Yane et al.. 2018a) (Hieh)
Acute
CD-I IGS outbred mice
(Mas musculus)
8-day LOEL
1,000
NA
Mortality
(Hazleton Laboratories.
1983)(Hieh)
Avian
Chronic
American kestrel (Fcdco
sparverius)
14-day LOEL
0.0025
NA
Thyroid
(Fernie et al.. 2015) (Hieh)
Soil invertebrates
Chronic
Earth worm (Eisenict
fetida)
3, 7, 14-day,
NOEC/LOEC
0.1/1.0
0.3
Gastrointestinal
(Yane et al.. 2018b) (Hieh)
Acute
Nematode
(Caenorhctbditis elegctns)
3-day NOEC/LOEC
6-day LC50
500/750
1,381
612.4
NA
Growth/mortality
(Xu et al.. 2017) (Hieh)
11 Hazard values for mammals and avian are in mg/kg-bw/day.
h Hazard value represented by the corresponding ""Endpoint" in the adjacent column (e.g., LC50, NOEC, LOEC)
c Geometric means of definitive values only (i.e.. >48 mg/kg was not used in the calculation).
Page 117 of 638
-------
4.2.4 Environmental Hazard Thresholds
EPA calculates hazard thresholds to identify potential concerns to aquatic and terrestrial species. For
aquatic species, the hazard threshold is called a concentration of concern (COC), and for terrestrial
species, the hazard threshold is called a hazard value or toxicity reference value (TRV). These terms
(COC, TRV, and hazard value) describe how the hazard thresholds are derived and can encompass
multiple taxa or ecologically relevant groups of taxa as the environmental risk characterization serves
populations of organisms within a wide diversity of environments. See Appendix G for more details
about how EPA weighed the scientific evidence. Hazard thresholds are then used to calculate RQs in the
risk characterization step of the environmental risk evaluation. After weighing the scientific evidence,
EPA selects the appropriate toxicity value from the integrated data to use as a hazard threshold for each
assessment type.
For aquatic species, EPA estimates hazard by calculating a COCs for a hazard threshold. COCs can be
calculated using a deterministic method by dividing a hazard value by an assessment factor (AF)
according to EPA methods (U.S. EPA 2016e. 2014b. 2012b).
Equation 4-1.
toxicity value
The AF approach is one-size-fits-all and does not take data availability or species sensitivity into
account (U.S. EPA, 2013, 2991006). COCs can also be calculated using probabilistic methods. For
example, an SSD can be used to calculate a hazardous concentration for 5 percent of species (HC05).
The HC05 estimates the concentration of TCEP that is expected to be protective for 95 percent of
species. This HC05 can then be used to derive a COC, and the lower bound of the 95 percent CI of the
HC05 can be used to account for uncertainty instead of dividing by an AF. The application of ICE
models eliminates the need for AFs by extrapolating toxicity to a diversity of species representing a
wide range of aquatic taxa with surrogate species sensitivity (Awkerman et al.. 2014). Aquatic hazard
values within Section 4.2.2 are presented in mg/L, while the subsequent section will demonstrate the
calculation of acute and chronic COC in |ig/L or ppb to conform with modeled and monitored
environmental media concentrations presenting within Section 4.3.
4.2.4.1 Aquatic Species COCs Using Empirical and SSD Data
For the acute COC, EPA used the 96-hour LC50 toxicity data from rainbow trout, zebrafish, and
Daphnia magna studies as surrogate species to predict LC50 toxicity values for 46 additional aquatic
organisms (22 fish, 1 amphibian, 9 aquatic invertebrates, 14 benthic invertebrate species) using the
Web-ICE application (Raimondo and Barron. 2010). Species with empirical hazard values (n = 7) and
predicted species (n = 46) toxicity data were subsequently used to calculate the distribution of species
sensitivity to TCEP exposure through the SSD toolbox as shown in Appendix G.2.1 (Etterson. 2020).
The calculated HC05 was 31.6 mg/L with a 95 percent CI of 16.7 mg/L to 57.0 mg/L (Figure_Apx G-4).
The lower 95 percent CI of the HC05 was then multiplied by 1,000 to convert mg/L to |ig/L (or ppb)
resulting in 16,700 |ig/L. The chronic COC was derived from the ChV of the 30-day LOEC/NOEC of
0.028 mg/L for yellow catfish with the application of an AF of 10. The ChV for yellow catfish
represents TCEP treatment mortality compared to control treatments observed within juveniles exposed
to TCEP for 30 days (Hu et al.. 2022). The ChV for growth and observations of liver histopathological
alterations identified from the same study from 30 day TCEP exposure is 0.0028 mg/L (Hu et al.. 2022).
For aquatic plants, Web-ICE and SSD were applied to five empirical and three estimated EC50 values as
detailed in Appendix G.2.1. Two model fit distributions (normal and logistic) demonstrated the best fit
Page 118 of 638
-------
and both resulted in a lower 95 percent CI of the EC50 of 66 mg/L. This value was then multiplied by
1,000 to convert mg/L to |ig/L (or ppb) resulting in 66,000 |ig/L.
The acute COC derived from the lower 95th percent confidence interval of the HC05 for TCEP is
16,700 |ig/L or ppb.
For the chronic COC, the ChV of the 30-day LOEC/NOEC of 0.028 mg/L for yellow catfish, based on
mortality. Therefore, the chronic COC = 0.028 mg/L/(AF of 10) = 0.0028 mg/L x 1,000 = 2.8 |ig/L or
ppb.
For the aquatic plants COC derived from the lower 95th percent confidence interval of the EC05 for
TCEP is 66,000 |ig/L or ppb.
4.2.4.2 Terrestrial Species Hazard Values
For terrestrial species, EPA estimates hazard by using a hazard value for soil invertebrates, a
deterministic approach, for calculating a TRV for mammals. The TRV is expressed as doses in units of
mg/kg-bw/day. Although the TRV for TCEP is derived from laboratory mice and rat studies, body
weight is normalized, therefore, the TRV can be used with ecologically relevant wildlife species to
evaluate chronic dietary exposure to TCEP. Representative wildlife species chronic hazard thresholds
were evaluated in the trophic transfer assessments using the TRV. The following criteria were used to
select the data to calculate the TRV with NOEL and/or LOEL data (U.S. EPA 2007a).
Step 1: At least three results and two species tested for reproduction, growth, or mortality general
end points.
• The minimum dataset required to derive either a mammalian or avian TRV consists of three
results (NOEL or LOEL values) for reproduction, growth, or mortality for at least two
mammalian or avian species. If these minimum results are not available, then a TRV is not
derived.
Step 2: Are there three or more NOELs in reproduction or growth effect groups?
• Calculation of a geometric mean requires at least three NOEL results from either the
reproduction or growth effect groups.
• Because there was a single reproduction effect result and no growth effect results, then
proceed to Step 3.
Step 3: If there is at least one NOEL result for the reproduction or growth effect groups:
• Then the TRV is equal to the lowest reported no-observed-adverse-effect level (NOAEL) for
any effect group (reproduction, growth, or mortality), except in cases where, the NOEL is
higher than the lowest bounded LOEL.
• Then the TRV is equal to the highest bounded NOEL below the lowest bounded LOEL.
For TCEP, the NOEL for reproduction is 350 mg/kg-bw/day, and the lowest mortality LOEL is 88
mg/kg-bw/day with a NOEL of 44 mg/kg-bw/day. For more details see Appendix G.2.2.
Toxicity Reference Value (TRV) for Terrestrial Toxicity
The chronic TRV for mammals is 44 mg/kg-bw/day.
For soil invertebrates, EPA estimates hazard by calculating the ChV for a hazard threshold. The ChV is
the geometric mean of the NOEC and LOEC values. Although the most sensitive adverse outcome from
Page 119 of 638
-------
TCEP exposure is for earthworm gastrointestinal damage, the ecologically relevant effects for soil
invertebrates are for reproduction, population, and growth. The nematode NOEC (500 mg/kg soil) and
LOEC (750 mg/kg soil) for reduced growth and shortened lifespan are used to calculate the ChV. The
ChV for soil invertebrates is 612.4 mg/kg soil.
^4.2,5 Summary of Environmental Hazard Assessment
Overall, EPA has moderate confidence in the evidence that TCEP presents hazard potential to aquatic
species (Table 4-8). For acute aquatic exposures to TCEP within vertebrates, the 96-hour LC50 toxicity
values are 46, 249.0, and 279.1 mg/L for Japanese seabass, rainbow trout and zebrafish, respectively,
from three high-quality studies (Zhang et al.. 2024; Alzualde et al.. 2018; Life Sciences Research Ltd.
1990a). Additional acute TCEP exposure within invertebrates resulted in empirical data for four species:
brine shrimp (Artemia sp.), Mysid shrimp (Neomysis awatschensis), Manila clam (Ruditapes
philippinarum), and Daphnia magna (Zhang et al.. 2024; Torav Research Center. 1997a). Empirical
acute hazard data from TCEP exposure is available from three fish and four invertebrate species and
when used with the Web-ICE application resulted in predicted species (n = 46) to calculate the
distribution of species sensitivity to acute TCEP exposure (Appendix G).
Subchronic exposures from one study and chronic exposure effects from five studies within aquatic
vertebrates represent a variety of exposure durations, life history stages, and fish species (Table 4-2).
Subchronic and chronic duration exposures of TCEP to aquatic vertebrates produced effects including:
mortality (Hu et al.. 2022; Wang et al.. 2022); growth/development (Hu et al.. 2023; Hu et al.. 2022;
Wang et al.. 2022; Sun et al.. 2016); organ level effects within liver (Hu et al.. 2023; Hu et al.. 2022;
Sutha et al.. 2020). gills (Sutha et al.. 2020). kidney (Sutha et al.. 2020). and heart rate (Wang et al..
2022). Chronic TCEP exposure in Daphnia magna produced effects on both mortality with a 21-day
LC50 of 83.1 mg/L and reproduction with a 21-day ChV of 13 mg/L (Torav Research Center. 1997c).
EPA identified two studies with overall quality determinations of high for green algae, resulting in five
hazard values for diatoms and green algae (Zhang et al.. 2024; Torav Research Center. 1997b).
Although no amphibian studies were available to assess potential hazards from TCEP exposure,
modeled data from Web-ICE provided acute hazard values from two surrogate species for bullfrog
resulting in a geometric mean LC50 of 264 mg/L.
EPA calculated COCs for aquatic organisms, which are summarized in Table 4-6. These COCs were
utilized to determine risk to aquatic organisms from modeled and published concentrations of TCEP in
surface water and benthic pore water. EPA calculated an acute COC from the lower 95 percent CI of the
HC05 of 16,7500 ppb for aquatic organisms based on the LC50 toxicity values from 3 test species and
19 fish, 1 amphibian, 7 aquatic invertebrates, and 10 benthic invertebrate species using Web-ICE
(Raimondo and Barron. 2010). Empirical and predicted species acute hazard data consisted of 25 fish, 1
amphibian, 12 aquatic invertebrates, and 15 benthic invertebrates used to calculate the distribution of
species sensitivity to TCEP exposure through the SSD toolbox (Etterson. 2020). The calculated HC05
was 31,600 |ig/L. The acute COC equals the lower 95 percent CI of the HC05, or 16,700 |ig/L. For the
chronic COC, the ChV of the 30-day LOEC/NOEC of 0.028 mg/L for yellow catfish based on mortality
(Hu et al.. 2022). was used with the application of an AF of 10, resulting in 2.8 ppb. The ChV for
growth and observations of liver histopathological alterations identified from the same study from 30
day TCEP exposure is 0.0028 mg/L (Hu et al.. 2022). The landscape of hazard values for aquatic plants
was largely represented within marine species with the resulting Web-ICE and SSD analysis utilizing
empirical and predicted species hazard values from two freshwater species and six saltwater species.
Two of the best fitting models both resulted in lower 95 percent CI value of the HC05, resulting in
66,000 |ig/L.
Page 120 of 638
-------
Overall, EPA has robust confidence in the evidence that TCEP presents hazard to terrestrial mammals
via dietary exposure, and moderate confidence that TCEP poses hazard to soil invertebrates (Table 4-7).
For chronic terrestrial mammalian exposures to TCEP, the NOEL, and/or LOEL toxicity data ranged
from a rat NOEL of 50 mg/kg-bw/day to a mouse LOEL of 1,000 mg/kg-bw/day for reproduction,
mortality, and/or neurotoxicity endpoints. These were assigned an overall quality determination of high
for all five studies with the exception of one medium overall quality determination for a reproduction
endpoint (Yang et al.. 2018a; Matthews et al.. 1993; NTP. 1991b; Matthews et al.. 1990; Hazleton
Laboratories. 1983). EPA calculated chronic toxicity to mammals from TCEP exposure using a TRV.
The TRV is equal to the highest NOAEL below the lowest LOAEL for mortality. The chronic TRV for
mammals is 44 mg/kg-bw/day (Table 4-7). The TRV is used as the chronic hazard threshold for
representative species during the trophic transfer assessments.
For soil invertebrate exposure to TCEP, a NOEC of 500 mg/kg soil and a LOEC of 750 mg/kg soil at 3
days exposure to TCEP was expressed for reduced growth and shortened lifespan of nematodes. The
ChV is 612 mg/kg soil for growth and reduced lifespan (Xu et al.. 2017) (Table 4-7).
Hazard threshold values for earthworms and American kestrels (Table 4-7) are represented by toxicity
endpoints, including degradation of the digestive track in earthworms and increases in plasma thyroid
hormones in kestrels. Although the most sensitive adverse outcome within soil invertebrates from TCEP
exposure is for earthworm, the ecologically relevant effects for soil invertebrates are for reduced growth
and shortened lifespan with a ChV of 612 mg/kg soil, from which an RQ value can be calculated.
Similarly, while the hazard value for the American kestrel within this analysis is based on elevated
plasma free thyroid concentrations at 7 days, the study did not detect any effects on free thyroid
concentrations, kestrel growth (i.e., body weight), nor food consumption at the conclusion of the 21-day
dietary exposure study with TCEP (Fernie et al.. 2015). Because the apical assessment endpoint of
growth was not affected, it is difficult to assess the ecological relevancy of the change.
Table 4-6. Environmental Hazard Thresholds for Aquatic Environmental Toxicity
Environmental Aquatic Toxicity
Hazard Value
(^g/L)
Assessment
Factor (AF)
COC
(^g/L)
Acute aquatic exposure:
Lower 95% CI of HC05 from SSD
16,700
N/A17
16,700
Chronic aquatic exposure: based on fish ChV
28
10
2.8
Aquatic Plants COC:
Lower 95% CI of HC05 from SSD
66,000
N/A17
66,000
17 Used lower 95% CI of the HC05 or EC05 to account for uncertainties rather than an AF.
Table 4-7. Environmental Hazard Thresholds for Terrestrial Environmental Toxicity
Environmental Terrestrial Toxicity
Hazard Value or TRV
Mammal
44 mg/kg-bw/day
American Kestrel (Fcdco sparverius)
0.0025 mg/kg-bw/day
Nematode (Caenorhctbditis elegans)
612 mg/kg soil
Earthworm (Eisenici fetidci)
0.3 mg/kg soil
Page 121 of 638
-------
4.2.6 Weight of Scientific Evidence Conclusions for Environmental Hazards
EPA uses several considerations when weighing and weighting the scientific evidence to determine
confidence in the environmental hazard data. These considerations include the quality of the database,
consistency, strength and precision, biological gradient/dose response, and relevance (see Appendix
G.2.3.1) and are consistent with the 2021 Draft Systematic Review Protocol (U.S. EPA 2021a). Table
4-8 summarizes how these considerations were determined for each environmental hazard threshold.
Overall, EPA considers the evidence for chronic mammalian hazard thresholds robust, the evidence for
aquatic vertebrate and invertebrate and terrestrial invertebrates hazard thresholds moderate, and the
evidence for chronic avian hazard thresholds slight. A more detailed explanation of the weight of
scientific evidence, uncertainties, and overall confidence levels is presented in Appendix G.2.3.1.
4.2.6.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the
Environmental Hazard Assessment
Quality of the Database; and Strength (Effect Magnitude) and Precision
All the studies used to calculate COCs (aquatic fish and algae), TRVs (terrestrial mammals), and hazard
thresholds (terrestrial invertebrates) received a high overall quality determination from the systematic
review data quality evaluation. Effect magnitude was not reported for mammal studies. Effect
magnitude was reported for aquatic fish, invertebrates, algae, and nematode studies using LC50s.
For aquatic organisms, six fish species, four invertebrate species, and five algal species were represented
in the empirical data from systematic review; seven invertebrate and vertebrate species had data
appropriate for the SSD model and all five algal species were appropriate for an SSD on aquatic plants.
EPA was able to supplement the dataset for 46 aquatic species for TCEP with predictions from Web-
ICE, which included predictions for 22 fish, 1 amphibian, 9 aquatic invertebrates, 14 benthic
invertebrates. The use of three surrogate species available as inputs for the Web-ICE application reduces
the confidence in the Web-ICE and subsequent SSD output. However, the use of the probabilistic
approach within this risk evaluation increases confidence compared to a deterministic approach using
the two studies on fishes and one study on Daphnia magna with acute hazard study endpoints. The use
of the lower 95 percent CI instead of a fixed AF of 5 also increases confidence as it is a more data-
driven way to account for uncertainty.
A total of seven studies represent subchronic and chronic exposure of TCEP to aquatic organisms and
were assigned moderate confidence in the overall quality of the database. The one chronic study on
invertebrates is a 21-day TCEP exposure to Daphnia magna reported effects on reproduction and
survival (Torav Research Center. 1997c). The remaining six studies are from TCEP exposures to fishes
(Japanese medaka, mrigal carp, yellow catfish, and zebrafish) for varying durations (14, 21, 30, and 120
days) and among different lifestages such as embryo/larvae (Wang et al.. 2022; Sun et al.. 2016).
juvenile (Hu et al.. 2022; Zhao et al.. 2021; Sutha et al.. 2020). and adult (Hu et al.. 2023). Several of the
chronic duration TCEP exposure studies recorded outcomes for apical assessment endpoints (i.e.,
survival, growth, reproduction) and sub-organ level assessment endpoints (i.e., hormone concentrations,
gene expression, oxidative stress). A study with a 30-day TCEP exposure with a ChV as an endpoint of
mortality was used to calculate the chronic COC (Hu et al.. 2022). The 30-day exposure was conducted
with analytical verification of TCEP concentrations throughout the duration of the exposure and was
accompanied by recording outcomes associated with effects on survival and growth in addition to and
mechanistic impacts on endocrine function and gene expression (Hu et al.. 2022). There were no
reasonably available empirical toxicity data available for benthic organisms. Using the acute and chronic
COCs creates an additional uncertainty associated with extrapolating water column organism sensitivity
from TCEP exposure.
Page 122 of 638
-------
The database for the aquatic plant assessment consisted of two studies on three species of green algae
and two diatom species both with an overall quality determination of high (Zhang et al.. 2024; Torav
Research Center. 1997b). Slight confidence was assigned to the overall quality of the database due to the
relatively limited number of studies and species represented. The previously conducted ECHA risk
assessment on TCEP summarizes hazards for two species of algae (Scenedesmus subspicatus and
Selencistrum capri cornutum) represented with five studies and reported a wide range of hazard values
for 48 to 96 hour exposures within algae (ECB. 2009). Moderate confidence was assigned to the strength
as precision consideration for the algal assessment with the use of analytically verified TCEP
concentrations and thorough protocols employed within the hazard study (Torav Research Center.
1997b).
For terrestrial mammal species, no wildlife studies were available from systematic review; however,
four high-quality level studies with two species, mice and rats, represented were used from human
health animal model studies. A TRV derived from the mammal studies was used to calculate the hazard
threshold in mg/kg-bw.
For avian species, a single, high-quality level study was available for the American kestrel. The avian
study detected transient differences in thyroid hormone level with no apparent effects on body weight or
food consumption. Although the test did not detect any effects on apical assessment endpoints of
regulatory interest (i.e., impaired growth, survival, or reproduction) and the ecological relevancy of
change in thyroid hormone level is uncertain, the study is still useful for the trophic transfer assessment.
For example, if the results of the trophic transfer show that exposure from TCEP is lower than (i.e., is
protective for) the hazard threshold for effect on thyroid hormones, then a qualitative assertion can be
made that the exposure levels from TCEP do not indicate risk.
For soil invertebrates, two high-quality level soil invertebrate studies were available. The earthworm
study did not have an ecologically relevant endpoint effect, although the earthworm is still useful for
assessing trophic transfer hazards both because of its direct ingestion of soil and because the earthworm
is expected to be part of the diet of other trophic levels (short-tailed shrew, woodcock, and American
kestrel).
Consistency: For aquatic fish species, the behavior effect of hypoactivity under dark phase stimulation
and development/growth effects was similar in Japanese medaka and zebrafish. Behavioral differences
between treatment and control groups were observed within mrigala carp from chronic 21-day exposures
to TCEP characterized by behaviors such as fast swimming and inability to feed (Sutha et al.. 2020).
Activity under light and dark phases, as well as development/growth effects, were not tested with
rainbow trout. Mortality effects for NOEC/LOEC and LC50s were similar for zebrafish and rainbow
trout while the LC50 for the Japanese seabass was lower indicating more sensitivity to TCEP. Many
chronic duration studies conducted aquatic vertebrates recorded histology, oxidative stress, gene
expression, endocrine function, and apical outcomes including but not limited to growth and survival.
Moderate confidence was assigned to the consistency of both acute and chronic exposure TCEP effects
on aquatic species because the outcomes were consistently observed within aquatic species.
For terrestrial mammal species, human health animal model studies (rats) are in agreement with respect
to neurotoxicity effects resulting from lesions to the brain. Confidence is robust on the MOA for rats on
exposure to TCEP via diet due to neurotoxic effects with lesions to the brain. Three studies included
mice; however, only a single study resulted in a LOEL for mortality. The maximum dose in all the
studies that included both rats and mice were all below the single study for mice where the lowest test
concentration resulted in the LOEL.
Page 123 of 638
-------
The single aquatic plant, avian, earthworm, and nematode studies were insufficient to characterize
consistency in their respective outcomes.
Biological Gradient/Dose-Response
A dose response was reported for all studies used for calculating hazard thresholds as well as the
earthworm study used in trophic transfer. However, because the American kestrel study only had one
dose concentration, no dose-response was reported.
Biological Relevance: Behavior and developmental/growth effects were in agreement between species
tested, zebrafish, yellow carp, mrigala carp and Japanese medaka (Hu et al.. 2023; Hu et al.. 2022; Sutha
et al.. 2020; Alzualde et al.. 2018; Sun et al.. 2016). Mortality effects from acute TCEP exposure were
also in agreement between species tested (Daphnia magna, brine shrimp, Manila clam, Japanese
seabass, mysid shrimp, zebrafish, and rainbow trout). All rat studies across multiple strains exhibited
brain lesions from TCEP exposure that was associated with the mortality endpoint. Data were
insufficient to observe correspondence of adverse outcomes across species within taxa group for avian
of terrestrial invertebrates.
Physical Chemical Relevance: Empirical data were on the effects of the chemical of interest, which
increases confidence. TCEP was identified, including source, for all organisms. Purity was either not
reported or not analytically verified for rainbow trout, earthworm, one of the mouse/rat studies
(Matthews et al.. 1990). and the American kestrel study (Fernie et al.. 2015).
Environmental Relevance: Additional uncertainty is associated with laboratory to field variation in
exposures to TCEP are likely to have some effect on hazard threshold; that is, gavage vs. natural forage
diet for mammals (rats and mice) and invertebrate substrate (i.e., nematodes maintained on nematode
growth medium and earth worms on artificial soil). The five species of green algae and diatoms
representing aquatic plant hazard were represented by four saltwater species (Zhang et al.. 2024) and
one freshwater species (Torav Research Center. 1997b). Test conditions for aquatic species correspond
well with natural environmental conditions.
Page 124 of 638
-------
Table 4-8. TCEP Evidence Tab
e Summarizing the Overall Confidence Derived
'rom Hazard Thresholds
Types of Evidence
Quality of the
Database
Consistency
Strength and
Precision
Biological Gradient/
Dose-Response
Relevance"
Hazard
Confidence6
Aquatic
Acute aquatic assessment
++
++
++
+++
+++
Moderate
Chronic aquatic assessment
++
++
++
+++
+++
Moderate
Algal assessment
+
Not applicable
++
++
+
Slight
Terrestrial
Chronic avian assessment
+
Not applicable
+
+
++
Slight
Chronic mammalian assessment
++
+++
+++
+++
+++
Robust
Terrestrial invertebrates
++
Not applicable
++
++
+++
Moderate
11 Relevance includes biological, physical/chemical, and environmental relevance.
b Hazard Confidence reflects the overall confidence in the conclusions about the presence or absence of hazard thresholds and the weight of support and
uncertainties around all the available data and does not necessarily represent a summation of the individual evidence properties.
+++ Robust confidence suggests thorough understanding of the scientific evidence and uncertainties. The supporting weight of scientific evidence outweighs the
uncertainties to the point where it is unlikely that the uncertainties could have a significant effect on the hazard estimate.
++ Moderate confidence suggests some understanding of the scientific evidence and uncertainties. The supporting scientific evidence weighed against the
uncertainties is reasonably adequate to characterize hazard estimates.
+ Slight confidence is assigned when the weight of scientific evidence may not be adequate to characterize the scenario, and when the assessor is making the best
scientific assessment possible in the absence of complete information. There are additional uncertainties that may need to be considered.
Page 125 of 638
-------
4.3 Environmental Risk Characterization
Environmental Risk Characterization (Section 4.3):
Key Points
EPA evaluated the reasonably available information to support environmental risk characterization.
The key points of the environmental risk characterization are summarized below:
• For aquatic species, the 5 of 20 COUs resulted in chronic RQs greater than 1 with over 30
days of exceedance within surface water (Table 4-21).
• For aquatic species, chronic RQs are above 1 and have corresponding days of exceedance
greater than 30 days within the sediment compartment (benthic pore water) for 5 of 20 COUs
(Table 4-21). Because of TCEP's affinity to bind to sediment and persistence in the aquatic
compartment, there could be a lasting effect on benthic biota and potential community-level
impacts from chronic TCEP exposure. EPA has moderate confidence in the RQ inputs for the
acute and chronic aquatic assessment.
• Monitoring data show RQs from mean TCEP surface water concentrations within the WQP
database or published literature below 1 (Table 4-13). However, differences in magnitude
between modeled and measured concentrations may be due to measured concentrations not
being geographically or temporally close to releases of TCEP from a facility.
• For terrestrial species, EPA did not identify RQs greater than or equal to 1.
o RQs for soil invertebrates or terrestrial mammals were less than 1 using either modeled
soil concentrations or concentrations taken from the very limited monitoring data set
available (from an urban area of Germany) (Table 4-22). EPA has moderate confidence in
the RQ inputs for the terrestrial invertebrate assessment.
o RQs were below 1 for all representative species and corresponding trophic level using
TCEP soil concentrations from available published literature. RQs were below 1 for
semi-aquatic terrestrial receptors via trophic transfer from fish and using the highest
modeled TCEP surface water concentrations (Table 4-22). EPA has moderate confidence
in the RQ inputs for the screening level trophic transfer assessment.
EPA considered fate, exposure, and environmental hazard to characterize the environmental risk of
TCEP. For environmental receptors, EPA estimated (1) risks to aquatic species via water and sediment,
and (2) to terrestrial species via exposure to soil by air deposition and through diet via trophic transfer.
Risk estimates to aquatic-dependent terrestrial species included exposures to TCEP through water and
diet. As described in Section 2.2.2, TCEP is described as a "ubiquitous" contaminant because it is
commonly found in various environmental compartments such as surface water, soil, sediment, and
biota. TCEP's physical and chemical properties suggests that its main mode of distribution in the
environment is water and soil, depending on the media of release (Figure 2-1; see Appendix F.2.1.2).
TCEP has the potential to undergo long-range transport in air and water (LTRP) that could be
significantly underestimated when using its physical and chemical properties in QSAR models. TCEP's
behavior in the environment often does not align with its physical and chemical properties. It can be
transported to sediment from overlying surface water by advection and dispersion of dissolved TCEP
and by deposition of suspended solids containing TCEP. However, TCEP can partition between surface
water and sediments due to its log Koc values (3.23 to 3.46) (Zhang et al.. 2021; Wang et al.. 2018a;
Zhang et al.. 2018b) and water solubility (7,820 mg/L) (U.S. EPA 2015b; EC. 2009; ECB. 2009). which
could contribute to its mobility in the environment. For example, TCEP in the soil was seen to be
Page 126 of 638
-------
vertically transported to deeper soil horizons, causing TCEP concentrations in the surface soil to be
lower (He et al.. 2017; Bacaloni et al.. 2008). TCEP does not undergo hydrolysis under environmentally
relevant conditions and is considered persistent in water (see Appendix F.2.3.1), sediment (see
Appendix F.2.3.2), and soil (see Appendix F.2.4.1).
Direct exposure of TCEP to terrestrial receptors via air was not assessed quantitatively because dietary
exposure was determined to be the driver of exposure to wildlife (Section 4.1.3). The contribution of
exposure risk from inhalation relative to the ingestion exposure route is not expected to drive risk
because of dilution-associated environmental conditions (U.S. EPA 2003a. b). The gaseous phase of
TCEP is expected to have a short half-life in the atmosphere (ti/2 = 5.8 hours) with a log Koa value of
7.86 to 7.93 (Okeme et al.. 2020). suggesting this compound would adsorb to organic carbon present in
airborne particles (Okeme et al.. 2020; Ji et al.. 2019; U.S. EPA. 2017a; Wang et al.. 2017b). The
resulting particle-bound TCEP would be expected to be removed from the atmosphere through wet or
dry deposition. Annual air deposition to water and soil was modeled using AERMOD for applicable
COUs (see Table 4-9), and these modeled values are included as components within the current
environmental risk characterization.
EPA quantitatively assessed TCEP concentrations in surface water, pore water, sediment, and soil for
aquatic and terrestrial receptors via modeled concentrations (EFAST 2014, VVWM-PSC, AERMOD)
representing COU-based releases of TCEP. As reported in Section 3.3.2.5, EPA estimated surface water
concentrations from COU based releases of TCEP and reported from 2,350 ppb (or |ig/L) to 10,000 ppb
with a production volume of 2,500 lb/year using high-end 95th percentile estimates. Considered to be a
minor component, annual air deposition of TCEP to water was modeled using AERMOD indicating
deposition to a lentic (i.e., relatively static) system at 1,000 m from the source at 0.49 ppb with a
production volume of 2,500 lb/year using high-end 95th percentile estimates, which is four orders of
magnitude less than the lowest surface water concentration modeled using the model, VVWM-PSC.
median TCEP surface water concentrations in ambient water were 0.23 ppb and ranged from 0.47 ppb to
7.66 ppb for 466 detected values in the WQP from 2003 to 2022. TCEP water concentrations in
published literature were reported in Section 3.3.2 and represent ambient TCEP concentrations from
surface waters and are not associated with direct environmental releases of TCEP. Maximum TCEP
concentrations in surface waters were collected near urban environments recorded at 0.581, 0.785, and
0.810 ppb during low-flow conditions in the Los Angeles, San Gabriel, and Santa Clara Rivers in
California, respectively (Maruva et al.. 2016; Sengupta et al.. 2014).
As reported in Section 3.3.2.9, modeled benthic pore water TCEP concentrations ranged from 90 to 334
ppb for the production volume of 2,500 lb/year. Modeled sediment concentrations ranged from 893 ppb
(or |ig/kg) to 1,960 ppb for the production volume of 2,500 lb/year. Air deposition to sediment, as
reported in Section 3.3.2.10, indicated the highest annual deposition at 1,000 m was 125 ppb, which is
almost 7 times lower than the lowest sediment TCEP value modeled with VVWM-PSC (Incorporation
into paints and coatings - solvent borne at 893 ppb) and about 40 times lower than the highest PSC
value for laboratory chemicals (5,040 ppb). As reported in Section 3.3.3.2, calculated TCEP soil
concentrations resulting from modeled air deposition 1,000 m from the source with a production volume
of 2,500 lb/year ranged from 1.49xl0~6 to 0.0039 mg/kg and 1.92xl0~6 to 0.0055 mg/kg for central
tendency and high-end meteorology conditions.
Section 4.2 details available environmental hazard data and indicates that TCEP presents hazard to
aquatic and terrestrial organisms. For acute exposures, TCEP is a hazard to aquatic animals at 16,700
ppb based on the lower 95 percent CI of the HC05 resulting from an SSD utilizing EPA's Web-ICE
(Raimondo et al.. 2023) and SSD toolbox applications (Etterson. 2020). For chronic exposures, TCEP is
Page 127 of 638
-------
a hazard to aquatic organisms with a ChV of 2.8 ppb for fish. For terrestrial exposures, TCEP is a hazard
to mammals at 44 mg/kg-bw/day and a hazard to soil invertebrates with a ChV of 612 mg/kg. In
addition, TCEP presented sub-organ level hazard values for birds at doses of 0.0025 mg/kg-bw/day and
for soil invertebrates at 0.3 mg/kg soil and will serve to supplement terrestrial receptors via a
conservative approach to estimate risk from trophic transfer.
EPA assigned an overall quality determination of high or medium to 23 acceptable aquatic toxicity
studies and 17 acceptable terrestrial toxicity studies (see Risk Evaluation for Tris(2-chloroethyl)
Phosphate - Systematic Review Supplemental File: Data Quality Evaluation of Environmental Hazard
Studies (U.S. EPA. 2024wV). The Risk Evaluation for Tris(2-chloroethyl) Phosphate - Systematic
Review Supplemental File: Data Quality Evaluation of Environmental Hazard Studies (U.S. EPA.
2024w) presents details of the data evaluations for each study, including evaluations of each metric and
overall study quality level. As detailed in Section 4.2.6, EPA considers the evidence for terrestrial
chronic mammalian robust, the evidence for aquatic hazard thresholds and terrestrial invertebrates
moderate, and the evidence for the algae and terrestrial chronic avian slight.
4.3.1 Risk Characterization Approach
EPA characterized the environmental risk of TCEP using RQs (U.S. EPA. 1998b; Barnthouse et al..
1982). which are defined as follows:
Equation 4-2.
Environmental Exposure Concentration
RQ =
Hazard Threshold
Environmental exposure concentrations for each compartment (i.e., surface water, pore water, sediment,
and soil) were based on measured (i.e., monitored data and/or reasonably available literature) and/or
modeled (i.e., E-FAST 2014, VVMW-PSC, AERMOD) concentrations of TCEP from Section 3.3. EPA
calculates hazard thresholds to identify potential concerns to aquatic and terrestrial species. These terms
describe how the values are derived and can encompass multiple taxa or ecologically relevant groups of
taxa as the environmental risk characterization serves populations of organisms within a wide diversity
of environments. For hazard thresholds, EPA used the COCs calculated for aquatic organisms, and the
hazard values or TRVs calculated for terrestrial organisms as detailed within Section 4.2.
RQs equal to 1 indicate that environmental exposures are the same as the hazard threshold. If the RQ is
above 1, the exposure is greater than the hazard threshold. If the RQ is below 1, the exposure is less than
the hazard threshold. RQs derived from modeled data for TCEP are shown in Table 4-11 and Table 4-12
for aquatic organisms, and Table 4-15 for terrestrial organisms. For aquatic species, acute risk is
indicated when the RQ is greater than or equal to 1 for acute exposures, or chronic risk is indicated with
a RQ greater than or equal to 1 with days of exceedance at or above 30 days for chronic exposures. The
chronic COC was derived from a 30-day exposure; therefore, the days of exceedance to demonstrate risk
reflects the exposure period for that hazard value. For terrestrial species, RQ values are calculated from
the hazard value for soil invertebrates (nematode) and TRV for mammals as detailed in Section 4.2.4,
and risk is indicated when the RQ greater than or equal to 1.
EPA used modeled (e.g., E-FAST 2014, VVWM/PSC, AERMOD) and measured (e.g., monitoring
information from peer-reviewed literature or relevant databases) data to characterize environmental
concentrations for TCEP and to calculate the RQ. Table 4-9 represents the COUs with relevant
environmental releases represented in the current risk characterization on aquatic and terrestrial
receptors. Exposure data are especially helpful to characterize exposures from facilities and/or COUs. In
Page 128 of 638
-------
the absence of facility-specific releases for TCEP, estimated releases were generated for a generic
facility for each COU with production volume scenarios set at 2,500 lb/year (Table 4-9). Exposure data
and corresponding RQ values produced with a production volume of 25,000 lb/year are presented within
Appendix H. Surface water monitoring data on TCEP from available databases such as the WQP and
published literature were used as additional approaches to characterize risk to aquatic receptors. The
purpose of using monitored data and published literature, when available, was to determine if
concentrations in the ambient environment exceeded the identified hazard benchmarks for aquatic and
terrestrial receptors while also providing support for or concurrence with modeled concentrations.
As described in Section 3.3.3.2, IIOAC and subsequently AERMOD were used to assess the estimated
release of TCEP via air deposition from specific exposure scenarios to soil (Table 4-9). Estimated
concentrations of TCEP that could be in soil via air deposition at the community level (1,000 m from the
source) exposure scenarios have been calculated.
Table 4-9. Risk Characterization to Corresponding Aquatic and Terrestrial Receptors Assessed
'or the Following COUs
RQ Values
RQ Values
COU (Life Cycle Stage/Category/Subcategory)
OES
Calculated for
Aquatic
Receptors"
Calculated for
Terrestrial
Receptors6
Manufacture/Import/Import
Repackaging
Yes
Yes
Processing/Incorporation into formulation,
mixture, or reaction product/Paint and coating
manufacturing
Incorporation into
paints and coatings -
1-part coatings
Yes
Yes
Processing/Incorporation into formulation,
mixture, or reaction product/Paint and coating
manufacturing
Incorporation into
paints and coatings -
2-part reactive
coatings
Yes
Yes
Processing/Incorporation into formulation,
Formulation of TCEP
mixture, or reaction product/Polymers used in
into 2-part reactive
Yes
Yes
aerospace equipment and products
resins
Processing/Incorporation into article/Aerospace
equipment and products and automotive articles
Processing into 2-part
resin article
N/A'#
Yes
and replacement parts containing TCEP
Processing/Recycling/Recycling
Recycling e-waste
EPA did not have sufficient data to
estimate these releases'7
Distribution in Commerce/Distribution in
Distribution in
Distribution in commerce1'
commerce
commerce
Industrial use/Other use/Aerospace equipment and
products and automotive articles and replacement
parts containing TCEP
Installing article
(containing 2-part
resin) for aerospace
applications
(electronic potting)
Releases expected to be negligiblec
Commercial use/ Other use/Aerospace equipment
and products and automotive articles and
replacement parts containing TCEP
Installing article
(containing 2-part
resin) for aerospace
applications
Releases expected to be negligiblec
Commercial use/Paints and coatings/Paints and
Use of paints and
coatings
coatings - Spray
application
Yes
Yes
Page 129 of 638
-------
COU (Life Cycle Stage/Category/Subcategory)
OES
RQ Values
Calculated for
Aquatic
Receptors"
RQ Values
Calculated for
Terrestrial
Receptors6
Commercial use/Laboratory chemicals/Laboratory
chemicals
Lab chemical - Use of
laboratory chemicals
Yes
Yes
Commercial use/Furnishing, cleaning, treatment
care products/Fabric and textile products
End of service life disposal
(Releases and exposures not
quantified)c
Commercial use/Furnishing, cleaning, treatment
care products/Foam seating and bedding products
End of service life disposal
(Releases and exposures not
quantified)c
Commercial use/construction, paint, electrical,
and metal products/Building/construction
materials - Insulation
End of service life disposal
(Releases and exposures not
quantified)c
Commercial use/Construction, paint, electrical,
and metal products/Building/construction
materials - Wood and engineered wood products
- Wood resin composites
End of service life disposal
(Releases and exposures not
quantified)c
Consumer use/Paints and coatings/Paints and
coatings, including those found on automotive
articles and replacement parts
No quantified environmental
releases from consumer uses6
Consumer use/Furnishing, cleaning, treatment
care products/Fabric and textile products
No quantified environmental
releases from consumer uses6
Consumer use/Furnishing, cleaning, treatment
care products/Foam seating and bedding products
No quantified environmental
releases from consumer uses6
Consumer use/Construction, paint, electrical, and
metal products/Building/construction materials -
Insulation
No quantified environmental
releases from consumer uses6
Consumer use/Construction, paint, electrical, and
metal products/Building/construction materials -
Wood and engineered wood products - Wood
resin composites
No quantified environmental
releases from consumer uses6
Disposal/Disposal/Disposal
Waste disposal (Landfill or
Incineration, covered in each
COU/OES as opposed to a separate
COU)c
11 RQ values calculated for aquatic receptors based on TCEP releases from wastewater, WQP database, and published
literature
b RQ values calculated for terrestrial receptors based on TCEP releases as fugitive air and stack air deposition to soil,
trophic transfer, and published literature
c Section 3.2 provides details on these OESs
d Distribution in Commerce of TCEP consists of the transportation associated with the moving of sealed containers of
TCEP or sealed packages of TCEP containing products (Section 3.1.1).
'' Section 5.1.2.2.5 details the lack of information to characterize exposures for disposal of consumer wastes
EPA used 110AC and AERMOD to estimate annual air deposition from hypothetical facility releases
and calculate resulting surface concentrations to a pond. Air deposition to surface water for the 2,500
lb/year production volume scenario resulted in the highest annual deposition at 1,000 m of 0.49 |ig/L
which is approximately 200 times lower than the lowest surface water TCEP value modeled with
VVWM-PSC (laboratory chemicals at 96 pg/kg). RQs for each relevant COU listed in Table 4-9 were
Page 130 of 638
-------
calculated for annual air deposition to surface water at 1,000 m and are presented within Appendix H for
both production volumes and meteorological conditions. All RQ values for the production volume
scenario of 2,500 lb/year were less than 1, with the highest RQ at 0.17 for TCEP use in paints and
coatings at job sites. The higher production volume scenario of 25,000 lb/year also resulted in RQs that
were less than 1 for all COUs expect the paints and coatings at job sites COU. The low production
volume scenario modeling used high-end estimates for at 95th percentile of the mean. RQs for the mean
(50th percentile) air to water deposition with the AERMOD for both meteorological models were below
1. It is not anticipated that air deposition to water will significantly contribute as TCEP concentrations
within the water column, pore water, and sediment.
Frequency and duration of exposure can affect the potential for adverse effects in aquatic receptors.
Within the aquatic environment, a two-tiered modeling approach was employed to predict surface water,
pore water, and sediment TCEP concentrations. If E-FAST 2014 predicted 7Q10 surface water
concentrations were greater than the chronic or acute COCs, the VVWM-PSC model was then used to
confirm whether the predicted surface water concentration days of exceedance as determined by the
acute COC and chronic COC. For TCEP, all six applicable OESs (Table 4-9) modeled in E-FAST 2014
produced chronic RQ values greater or equal to 1, prompting the use of VVWM-PSC for greater
ecological resolution on TCEP concentrations and days of exceedance within the water column and
benthic compartments. VVWM-PSC considers model inputs of physical and chemical properties of
TCEP (i.e., Kow, Koc, water column half-life, photolysis half-life, hydrolysis half-life, and benthic half-
life) allowing EPA to model predicted benthic pore water and sediment concentrations.
Environmental RQ values by exposure scenario with TCEP surface water concentrations (ppb) were
modeled by E-FAST 2014 and VVWM-PSC and are presented in Table 4-11. The max day average
concentrations produced by VVWM-PSC represent the maximum concentration (ppb) over a 1- or 30-
day average period corresponding with the acute or chronic COC used for the RQ estimate.
Environmental RQ values by exposure scenario for aquatic organisms with TCEP pore water
concentration modeled by VVWM-PSC are presented within Table 4-12. Scenarios and production
volume allow for the calculation of RQs and days of exceedance that for risk estimation to aquatic
organisms (scenarios with an acute RQ greater than or equal to 1, or a chronic RQ greater than or equal
to 1 and 30 days or more of exceedance for the chronic COC).
EPA considers the biological relevance of species that COCs or hazard values are based on when
integrating these values with the location of the surface water, pore water, and sediment concentration
data to produce RQs. Life-history and habitat of aquatic organisms influence the likelihood of exposure
above the hazard threshold in an aquatic environment. EPA has identified COC values associated with
aquatic hazard values and include acute COC and chronic COC. The acute COC for aquatic species is
the lower 95 percent CI of the HC05 of an SSD, a modeled probability distribution of toxicity values
from multiple taxa inhabiting the water column. The chronic COC is represented by a mortality endpoint
from 30-day exposures to TCEP within the water column. Calculated RQ values for pore water are
represented with acute and chronic COCs. The confidence in these RQ inputs were described as
moderate confidence determinations for the acute COC and chronic COC.
4.3.1.1 Risk Characterization Approach for Trophic Transfer
Trophic transfer is the process by which chemical contaminants can be taken up by organisms through
dietary and media exposures and transfer from one trophic level to another. Chemicals can be transferred
from contaminated media and diet to biological tissue and accumulate throughout an organisms' lifespan
(bioaccumulation) if they are not readily excreted or metabolized. Through dietary consumption of prey,
a chemical can subsequently be transferred from one trophic level to another. If biomagnification occurs,
Page 131 of 638
-------
higher trophic level predators will contain greater body burdens of a contaminant compared to lower
trophic level organisms.
EPA conducted screening-level approaches for aquatic and terrestrial risk estimation based on exposure
via trophic transfer using conservative assumptions for factors such as: area use factor, TCEP absorption
from diet, soil, and water. Section F.2.5 details persistence as this compound is expected to persist
within aquatic and terrestrial environments. Under laboratory conditions, mean whole body BCF for
juvenile Atlantic Salmon (Salmo solar) is reported as 0.34 L/kg wet weight for an experimental
exposure concentration of 1.0 mg/L (Arukwe et al.. 2018). TCEP is not considered bioaccumulative;
however, geometric mean concentrations within biota in Lake Erie have been reported at concentrations
of 35.6 ng/g lipid as reported by Guo et al. (2017b) in Section 4.1.2. Section 4.1 reports measured
concentrations of TCEP within biota with seven studies indicating TCEP concentrations within whole
fish and lipid (see Section 4.1.2.1), one study within a marine mammal (see Section 4.1.2.1), and two
studies with terrestrial organisms (see Section 4.1.3.1). A screening4evel analysis was conducted for
trophic transfer and formulation of RQ values from aquatic and terrestrial hazard values. If RQ values
were greater than or equal to 1, risk estimation based on potential trophic transfer of TCEP is indicated
from this screening4evel approach and further refined analysis is warranted. If an RQ value is less than
1, risk based on potential trophic transfer of TCEP is not indicated from screening-level approach and no
further assessment is necessary. The screening-level approach employs a combination of conservative
assumptions (i.e., conditions for several exposure factors included within Equation 4-3 below) and
utilization of the maximum values obtained from modeled and/or monitoring data from relevant
environmental compartments.
A secondary source of TCEP contamination in soil is from the application of biosolids. For this
screening analysis, the COU with the highest release estimates were modeled with methods described
within Section 3.3.3.5. Using BST, EPA estimated soil concentrations of 0.1412 mg/kg for the 2,500
lb/year production volume high-end estimate for the Incorporation into paints and coatings - 1-part
coatings OES, and 0.5293 mg/kg for the 25,000 lb/year production volume, central tendency estimate
for the Incorporation into paints and coatings - 2-part reactive coatings OES.
Following the basic equations as reported in Chapter 4 of the U.S. EPA Guidance for Developing
Ecological Soil Screening Levels (U.S. EPA. 2005a). wildlife receptors may be exposed to contaminants
in soil by two main pathways: incidental ingestion of soil while feeding, and ingestion of food items that
have become contaminated due to uptake from soil. The general equation used to estimate the risk from
exposure via these two pathways is provided below:
Soilj *PS* FIR * AFsy] + [£f=1 Bij *Pt* [FIR + WIR] * AFiy]) * AUF
HTj
Risk quotient for contaminant (j) (unitless)
Concentration of contaminant (j) in soil (mg/kg dry weight)
Number of different biota type (i) in diet
Concentration of contaminant (j) in biota type (i) (mg/kg dry weight)
Proportion of biota type (i) in diet
Food intake rate (kg of food [dry weight] per kg body weight per day)
Water intake rate (kg of water per kg body weight per day)
Equation 4-3.
RQj = —
Where:
RQ, =
Soilj =
N
Bij =
Pi
FIR =
WIR =
Page 132 of 638
-------
AF a =
Absorbed fraction of contaminant (j) from biota type (i) (for screening
purposes set equal to 1)
kVsj =
Absorbed fraction of contaminant (j) from soil (s) (for screening purposes set
equal to 1)
HTj =
Hazard Threshold (mg/kg-BW[wet weight]/day)
Ps
Proportion of total food intake that is soil (kg soil/kg food)
AUF =
Area use factor (for screening purposes set equal to 1)
Table 4-10. Terms and Values Used to Assess Potential Trophic Transfer of TCEP for Terrestrial
Risk Characterization
Earthworm
(Eisenici fetida)
Short-Tailed Shrew
American
American Kestrel
(Fctlco sparverius)
American
Term
(Blarina
breviccmda)
Woodcock
(Scolopax minor)
Mink
(Mustela vison)
Soilja
0.0055 mg/kgfe
TCEP
0.0055 mg/kgfe
TCEP
0.0055 mg/kgfe
TCEP
0.0055 mg/kgfe
TCEP
10.3 mg/Lc
TCEP
N
1
1
1
3
1
0.0055 mg/kg
TCEP (worm)
By
0.0055 mg/kgfe
TCEP (soil)
0.0055 mg/kg TCEP
(worm)
0.0055 mg/kg
TCEP (worm)
0.0046 mg/kg
TCEP (short-tailed
shrew)
3.71 mg/kg''
TCEP (Fish)
0.0057 mg/kg
TCEP (woodcock)
P
1
1
1
0.33
1
FIR
1
0.55e
0.77e
0.30'#
0.22e
WIR
1
0.223e
o.r
Dietary hydration
0.104e
AFy
1
1
1
1
1
AF,;
1
1
1
1
1
HTj
0.3 mg/kg -
soil/day
44 mg/kg-bw/day
N/A'
0.0025 mg
TCEP/kg-bw/day
44 mg
TCEP/kg-
bw/day
Ps
1
0.03g
0.164g
0.057g
1
AUF
1
1
1
1
1
11 TCEP concentration in surface water for Mink
h Highest soil concentration of TCEP obtained using AERMOD modeling (2,500 lb/year)
c Highest surface water concentration of TCEP obtained using VVWM-PSC modeling (2,500 lb/year)
d Highest fish concentration (mg/kg) calculated from surface water concentration TCEP (WWM-PSC) and whole
bodv BCF of 0.34 (Arukwe et al.. 2018)
'' Exposure factors (FIR and WIR) sourced from EPA's Wildlife Exposure Factors Handbook (U.S. EPA. 1993b)
' No TCEP hazard threshold value for this representative species is available
g Soil ingestion as proportion of diet represented at the 90th percentile sourced from EPA's Guidance for Developing
Ecological Soil Screening Levels (U.S. EPA. 2005a)
Terrestrial hazard data are available for soil invertebrate and mammals using hazard values detailed in
Section 4.2.4. Representative avian and mammal species are chosen to connect the TCEP transport
exposure pathway via trophic transfer from earthworm uptake of TCEP from contaminated soil through
invertivore avian (American woodcock) and mammal (short-tailed shrew) species, to the American
kestrel that feeds on invertebrates as well as avian and small terrestrial vertebrates.
Page 133 of 638
-------
At the screening-level, the conservative assumption is that the invertebrate diet for the American
woodcock and short-tailed shrew comprises 100 percent earthworms from contaminated soil. Similarly,
the dietary assumptions for the American kestrel are 100 percent of the invertebrate, avian, and mammal
diet are from the earthworm, American woodcock, and short-tailed shrew, respectively. Additionally,
the screening-level analysis uses the highest modeled or monitored soil contaminate level to determine if
a more detailed assessment is required. Because surface water sources for wildlife water ingestion are
typically ephemeral, the trophic transfer analysis for terrestrial organism assumed TCEP exposure
concentration for wildlife water intake are equal to soil concentrations for each corresponding exposure
scenario.
Exposure factors for food intake rate (FIR) and water intake rate (WIR) were sourced from the EPA's
Wildlife Exposure Factors Handbook (U.S. EPA. 1993b). The proportion of total food intake that is soil
(.Ps) is represented at the 90th percentile for representative taxa (short-tailed shrew, woodcock, and
hawk) and was sourced from calculations and modeling in EPA's Guidance for Developing Ecological
Soil Screening Levels (U.S. EPA. 2005a). Additional assumptions for this analysis have been considered
to represent conservative screening values (U.S. EPA. 2005a). Within this model, incidental oral soil
exposure is added to the dietary exposure resulting in total oral exposure greater than 100 percent. In
addition, EPA assumes that 100 percent of the contaminant is absorbed from both the soil (AF*,-) and
biota representing prey (AFy). The proportional representation of time an animal spends occupying an
exposed environment is known the area use factor (AUF) and has been set at 1 for all biota within this
equation (Table 4-10).
The following hazard values were used for trophic transfer of TCEP from media (soil) through trophic
levels: earthworm ChV of 0.3 mg/kg soil, mammal TRV dose of 44 mg/kg-bw/day, and American
kestrel LOEL at doses of 0.0025 mg/kg-bw/day. Short-tailed shew and American mink hazard threshold
values were calculated from the mammal TRV (44 mg/kg-bw/day). It is important to reiterate that
hazard values within this screening-level trophic transfer analysis for earthworm and American kestrel
are represented by endpoints of gastrointestinal damage and increased plasma thyroid hormones,
respectively. Although the most sensitive adverse outcome within soil invertebrates from TCEP
exposure is for earthworm, the ecologically relevant effects for soil invertebrates are for reduced growth
and shortened lifespan with a ChV of 612 soil mg/kg from which an RQ value can also be calculated.
The inclusion of earthworms and kestrels from this screening-level analysis represent an additional
conservative approach for estimating risk to terrestrial organisms via trophic transfer.
For semi-aquatic terrestrial species, the TRV was used with the American mink for the screening-level
assessment (Table 4-10). Similar to the above soil concentrations used as term Soil in Equation 4-1, the
highest surface water concentration modeled via VVWM-PSC was used as a surrogate for the TCEP
concentration found in the American mink's diet, which is highly variable depending on habitat. In a
riparian habitat, mink derive 74 to 92 percent of their diet from aquatic organisms, which includes fish,
crustaceans, birds, mammals, and vegetation (Alexander. 1977). The American mink was used as the
representative species for semi-aquatic mammals. As a conservative assumption, 100 percent of the
American mink's diet is predicted to come from fish. Fish concentration (mg/kg) was calculated using
surface water concentrations of TCEP from VVWM-PSC assuming a BCF of 0.34 as reported for whole
body values from 1 mg/L TCEP exposures under laboratory conditions (Arukwe et al.. 2018).
4.3.2 Risk Characterization for Aquatic Receptors
The VVWM-PSC model identified substantial deposition of TCEP to the sediment and resulting pore
water (Table 4-12) with a production volume of 2,500 lb/year. A major concern centered around the
RQs within pore water is the lasting effects on benthic biota and potential community-level impacts
Page 134 of 638
-------
from chronic TCEP exposure within this aquatic compartment. Acute risk estimates for both surface and
pore water were not at or above one for any COU, however, chronic risk estimates were greater than one
for 5 out of 20 COUs with quantified releases to water for both surface water (Table 4-11) and pore
water (Table 4-12) using high-end estimates for the 2,500 lb/year production volume scenario.
Furthermore, RQ values are also greater than one for both surface water (TableApx H-2) and pore
water (Table Apx H-3) the same 5 out of 20 COUs when using the central tendency estimates for the
2,500 lb/year production volume scenario.
The physical and chemical properties of TCEP and its persistence translate to removal from the water
column by particulate and sediment organic matter and persistence within sediment (see Section 2.2.2).
TCEP may partition between water and sediment due to its physical and chemical properties and, as a
result, exposure of TCEP and the duration of that exposure to organisms dwelling within the sediment
could be elevated. Many benthic invertebrates are detritivores, meaning they feed on dead plant and
animal material or contribute to the liberation of additional nutrient resources by further breaking down
these materials. Detritivorous benthic invertebrates often serve as an important food source for many
juvenile fishery and non-game resident species. In several cases, days of exceedance were greater in
pore water (Table 4-12) than the surface water (Table 4-11), further indicating that TCEP would be a
more persistent hazard to benthic dwelling organisms with increased durations of exposure.
Listed below are the 5 out of 20 COUs (Life Cycle Stage/Category/Subcategory with their respective
OES) evaluated, RQs for chronic duration exposures were greater than or equal to one with more than
30 days of exceedance within surface water and pore water.
Manufacture/Import/Import/Import and Repackaging
Surface Water: Surface water acute RQ values for import and packaging TCEP was less than 1 via both
E-FAST 2014 and VVWM-PSC modeling. VVWM-PSC demonstrated a chronic RQ greater than 1 at
112.5 with 34 days of exceedance.
Pore Water: The pore water acute RQ for importing and repackaging TCEP was less than one the acute
COC. The chronic RQ for importing and repackaging TCEP was greater than one for the chronic COC
at 44.3. The corresponding days of exceedance for the chronic COC was 252 days.
Processing/Incorporation into Formulation, Mixture, or Reaction Product/Paints and Coating
Manufacturing/Incorporation into Paints and Coatings - 1-Part Coatings
Surface Water: Surface water acute RQ values for TCEP incorporation into paints and coatings - 1-part
coatings were less than 1 via both E-FAST 2014 and VVWM-PSC modeling. VVWM-PSC
demonstrated a chronic RQ greater than 1 at 244.3 with 80 days of exceedance.
Pore Water: The pore water acute RQ for TCEP incorporation into paints and coatings - 1-part coatings
was less than one for the acute COC. The chronic RQ for importing and repackaging TCEP was greater
than one for the chronic COC at 96.8. The corresponding days of exceedance for the chronic COC was
298 days.
Processing/Incorporation into Formulation, Mixture, or Reaction Product/Paints and Coating
Manufacturing/Incorporation into Paints and Coatings - 2-Part Coatings
Surface Water: Surface water acute RQ values for TCEP incorporation into paints and coatings -
resins/solvent-borne were less than 1 via both E-FAST 2014 and VVWM-PSC modeling. VVWM-PSC
demonstrated a chronic RQ greater than 1 at 110 with 33 days of exceedance.
Page 135 of 638
-------
Pore Water: The pore water acute RQ for TCEP incorporation into paints and coatings - resins/solvent-
borne was less than one for the acute COC. The chronic RQ for importing and repackaging TCEP was
greater than one for the chronic COC at 43.9. The corresponding days of exceedance for the chronic
COC was 251 days.
Commercial use/Paints and coatings/Paints and coatings/Use in Paints and Coatings - Spray
application
Surface Water: Surface water acute RQ values for TCEP use in paints and coatings at job sites were less
than 1 via both E-FAST 2014 and VVWM-PSC modeling. VVWM-PSC demonstrated a chronic RQ
greater than 1 at 132.5 with 33 days of exceedance.
Pore Water: The pore water acute RQ for TCEP use in paints and coatings at job sites was less than one
for the acute COC. The chronic RQs for paints and coatings at job sites was greater than one for the
chronic COC at 52.5. The corresponding days of exceedance for the chronic COC was 260 days.
Processing/Incorporated into Formulation, Mixture, or Reaction Product/Polymers Used in
Aerospace Equipment and Products/Formulation of TCEP into 2-Part Reactive Resins
Surface Water: Surface water acute RQ values for formulation of TCEP into 2-part reactive resins were
less than 1 via both E-FAST 2014 and VVWM-PSC modeling. VVWM-PSC demonstrated a chronic
RQ greater than 1 at 129.3 with 52 days of exceedance.
Pore Water: The pore water acute RQ for formulation of TCEP into 2-part reactive resins was less than
one for the acute COC. The chronic RQ for 2-part reactive resins was greater than one for the chronic
COC at 51.4. The corresponding days of exceedance for the chronic COC was 262 days.
Commercial Use/Laboratory Chemicals/Laboratory Chemicals/Laboratory Chemicals
Surface Water: Within the water column, acute RQ values for laboratory chemicals were less than 1 via
both E-FAST 2014 and VVMM-PSC modeling. VVMW-PSC modeling demonstrated a chronic RQ
greater than 1 of 34.5 with days of exceedance of 210.
Pore Water: The pore water acute RQs for laboratory chemicals was less than one for the acute COC.
The chronic RQ for laboratory chemicals was greater than one at 32. The corresponding days of
exceedance for the chronic COC was 364 days.
Page 136 of 638
-------
Table 4-11. Environmental RQs by COU with Production Volumes of 2,500 lb/year for Aquatic Organisms with TCEP Surface Water
Concentration (ppb) Modeled by VVWM-PSC
COU (Life Cycle
Stage/Category/Subcategory)
OES
Production
Volume
(lb/year)fl
Days of
Release
Release
(kg/day)
Modeled Using VVWM-PSC
Max Day
Average
(ppb)6
COC
Type
COC
(PPb)
Days of
Exceedance
(days per
year)
RQ
Manufacture/Import/Import
Import and
repackaging
2,500
4
9.88
2,350
Acute
16,700
N/A
0.14
315
Chronic
2.8
34
112.5
Processing/Incorporation into
formulation, mixture, or reaction
product/ Paint and coating
manufacturing
Incorporation into
paints and coatings -
1-part coatings
2,500
2
35.17
10,000
Acute
16,700
N/A
0.60
684
Chronic
2.8
80
244.29
Processing/Incorporation into
formulation, mixture, or reaction
product/Paint and coating
manufacturing
Incorporation into
paints and coatings -
2-part reactive
coatings
2,500
1
31.89
8,150
Acute
16,700
N/A
0.49
310
Chronic
2.8
33
110.71
Commercial use/Paints and
coatings/Paints and coatings
Use in paints and
coatings - Spray
application
2,500
2
23.25
5,500
Acute
16,700
NA
0.33
371
Chronic
2.8
33
132.5
Processing/Incorporation into
formulation, mixture, or reaction
product/Polymers used in aerospace
equipment and products
Formulation of
TCEP into 2-part
reactive resins
2,500
1
31.53
9,040
Acute
16,700
N/A
0.54
362
Chronic
2.8
52
129.28
Commercial use/Laboratory
chemicals/ Laboratory chemicals
Laboratory
chemicals
2,500
182
0.39
96.6
Acute
16,700
N/A
5.80E-03
96.5
Chronic
2.8
210
34.46
N/A = Days of exceedance are modeled for the application of c
11 Production volume of 2,500 lb TCEP/year uses high-end estin
h Max day average represents the maximum concentration over
estimate.
ironic COCs and do not apply for acute COCs and corresponding RQs
lates (95th percentile for all COUs except the laboratory chemicals COU uses the 1st percentile),
a 1- or 30-day average period corresponding with the acute or chronic COC used for the RQ
Page 137 of 638
-------
Table 4-12. Environmental RQs by COU with Production Volumes of 2,500 lb/year for Aquatic Organisms with TCEP Pore Water
Concentration (ppb) Modeled by VVWM-PSC
COU (Life Cycle
Stage/Category/Subcategory)
OES
Production
Volume
(lb/year)fl
Days of
Release
Release
(kg/day)
Benthic Pore
Water
Concentration
(ppb)6
Benthic Pore Water
COC
Type
COC
(PPb)
Days of
Exceedance
RQ
Manufacture/Import/Import
Import and
repackaging
2,500
4
9.88
153
Acute
16,700
N/A
9.16E-03
124
Chronic
2.8
252
44.28
Processing/Incorporation into
formulation, mixture, or
reaction product/Paint and
coating manufacturing
Incorporation into
paints and coatings -
1-part coatings
2,500
2
35.18
334
Acute
16,700
N/A
0.02
271
Chronic
2.8
298
96.78
Processing/Incorporation into
formulation, mixture, or
reaction product/Paint and
coating manufacturing
Incorporation into
paints and coatings -
2-part reactive
coatings
2,500
1
31.89
152
Acute
16,700
N/A
9.10E-03
123
Chronic
2.8
251
43.92
Commercial use/Paints and
coatings/Paints and coatings
Use in paints and
coatings - Spray
application
2,500
2
23.26
182
Acute
16,700
N/A
1.09E-02
147
Chronic
2.8
260
52.5
Processing/Incorporation into
formulation, mixture, or
reaction product/Polymers
used in aerospace equipment
and products
Formulation of
TCEP into 2-part
reactive resins
2,500
1
31.53
177
Acute
16,700
N/A
1.06E-02
144
Chronic
2.8
262
51.4
Commercial use/Laboratory
chemicals/Laboratory
chemicals
Laboratory
chemicals
2,500
182
0.40
90.5
Acute
16,700
N/A
5.42E-03
89.6
Chronic
2.8
364
32
N/A = Days of Exceedance are modeled for the application of chronic COCs and c
o not apply for acute COCs and corresponding RQs
11 Production volume of 2,500 lb TCEP/year uses high-end estimates (95th percentile for all COUs except the laboratory chemicals COU uses the 1st
percentile).
Max day average represents the maximum concentration over a 1- or 30-day average period corresponding with the acute or chronic COC used for the RQ
estimate.
Page 138 of 638
-------
EPA used surface water monitoring data from the WQP and published literature to characterize the risk
of TCEP to aquatic organisms. These monitored surface water data reflect concentrations of TCEP in
ambient water. WQP data show an average (± SEM) concentration for TCEP of 0.33 ± 0.02 ppb in
surface water from 466 measurements taken throughout the United States between 2003 and 2022. The
highest concentration recorded during this period was 7.66 ppb, which was recorded in August 2013 in
Rochester, New York. Table 4-13 shows that RQ estimates were less than 1 for both acute and chronic
COCs using the mean value from the WQP, however, the maximum recorded TCEP concentration
within the WQP data show a RQ greater than 1.
Table 4-13. RQs Calculated Using Monitored Environmental Concentrations from WQX/WQP
Monitored Surface Water Concentrations
(ppb) from 2003-2022
RQ Using Acute COC of
16,700 ppb
RQ Using Chronic COC of
2.8 ppb
Mean (Standard Error of the Mean):
0.33 (0.02) ppb
1.97E-05
0.11
Maximum: 7.66 ppb
4.58E-04
2.73
Five of the six studies from reasonably available published literature sampled waters within the United
States, while one included sample sites from both U.S. and Canadian waters (Scott et al.. 1996). All six
studies from published literature are represented by general population surface water sampling where
TCEP concentration are not associated with a specific facility. One study encompassed 85 sample sites
for TCEP with study design placing sampling directly downstream from "intense urbanization and
livestock production, detecting TCEP within 49 of the 85 samples and resulting in minimum and
maximum TCEP concentrations of 0.02 and 0.54 ppb, respectively" (Kolpin et al.. 2002). Across all
studies a total of 185 samples resulted in 141 samples with TCEP detected and 44 non-detected
samplings between 1994 and 2013. The mean (±SEM) for TCEP concentrations reported within surface
water in the reasonably available published literature is 0.16 (±0.05) ppb with minimum and maximum
concentrations of 0.0002 and 0.81 ppb, respectively. Table 4-14 shows RQs all estimates are less than
one for both acute and chronic COCs.
Table 4-14. RQs Calculated Using TCEP in Surface Water from Monitored Environmental
Concentrations from Published Literature
Monitored Surface Water Concentrations
(ppb) from Published Literature
RQ Using Acute COC of
16,700 ppb
RQ Using Chronic COC of
2.8 ppb
Mean (Standard Error of the Mean):
0.16 (0.05) ppb
9.58E-06
5.71E-02
Maximum: 0.81 ppb
4.85E-05
0.29
4.3.3 Risk Characterization for Terrestrial Receptors
RQs were less than 1 for all relevant exposure scenarios when using the highest AERMOD predictions
for air deposition to soil at 1,000 m. Table 4-15 presents soil concentration and chronic RQ values from
the exposure scenario with the highest TCEP soil concentrations, indicating RQs below 1 for soil
organisms based on modeling data. The highest soil concentration recorded from AERMOD predictions
is 0.0055 mg/kg based on TCEP use in paints and coatings at job sites at 1,000 m. Soil concentrations
and RQ values for all scenarios, production volumes, and meteorology models are presented within
TableApx H-12.
Page 139 of 638
-------
Table 4-15. Calculated RQs Based on TCEP Soil Concentrations (mg/kg) as Calculated Using
Modeled Data
OES
Production
Volume
(lb/year)fl
Meteorological
Model6
Soil Concentration
(mg/kg) at 1,000 mc
Chronic RQ (Hazard
Value: 612 mg/kg)
Use in paints and
2,500
MetCT
3.97E-03
6.49E-06
coatings at job sites
MetHIGH
5.58E-03
9.11E-06
11 Production volume of 2,500 lb TCEP/yr uses high-end estimates (95th percentile).
h The ambient air modeled concentrations and deposition values are presented for two meteorology conditions
(Sioux Falls, South Dakota, for central tendency meteorology; and Lake Charles, Louisiana, for higher-end
meteorology).
c Estimated concentrations of TCEP (90th percentile) that could be in soil via air deposition at a community (1,000
m from the source) exposure scenario.
Risk characterization and trophic transfer for terrestrial receptors is based on modeled soil data from
AERMOD because there are no published literature or monitoring databases with TCEP soil
concentrations from U.S. sites and one comparative study from Germany (Mihailovic and Fries. 2012).
Transient increases in TCEP concentration have been observed with mean concentrations elevated from
0.008 to 0.023 mg/kg immediately following snowmelt conditions (Mihailovic and Fries. 2012). RQs to
soil invertebrates were below 1 for soil TCEP concentrations as reported for different sample periods
from Mihailovic and Fries (2012) (Table 4-16).
Table 4-16. RQs Calculated Using TCEP Soil Concern
trations from Published Literature
Sample Collection
Conditions
Mean TCEP
Concentration in Soil
(mg/kg)
Chronic RQ (Hazard
Value: 612 mg/kg)
Reference
(Overall Quality
Determination)
Soil TCEP concentrations in
January
5.89E-03
9.62E-06
(Mihailovic and
Fries. 2012) (Hish)
Soil TCEP concentration prior
to snowmelt
7.67E-03
1.25E-05
Soil TCEP concentration 24
hours after snowmelt
2.34E10-02
3.76E-05
4,3.4 Risk Characterization Based on Trophic Transfer in the Environment
Trophic transfer of TCEP and potential risk to terrestrial animals was evaluated using a screening-level
approach conducted as described in the EPA's Guidance for Developing Ecological Soil Screening
Levels (U.S. EPA. 2005a). TCEP concentrations within biota and resulting RQ values for all six relevant
COUs represented by seven OESs (Table 4-9), two production volume scenarios (2,500 and 25,000
lb/year), and two meteorological models for soil deposition are presented in Table Apx H-13.
Table 4-17 presents biota concentrations and RQ values for the highest soil concentration via AERMOD
(Paints and coatings) at the 2,500 production volumes. RQs were below 1 for all soil concentrations and
COUs based on the chronic hazard threshold for terrestrial invertebrate identified within Section 4.2.4.2.
The chronic TRV, calculated using empirical toxicity data with mice and rats, also resulted in RQs less
than 1 for all modeled soil concentrations. The overall hazard confidence for the chronic mammalian
assessment and terrestrial invertebrates reported within Section 4.2.6 as robust and moderate,
respectively, providing increased confidence in the application of these ecologically relevant hazard
thresholds.
Page 140 of 638
-------
Estimates of risk represented as RQ values were calculated using hazard thresholds with in vivo data
measuring ecologically relevant endpoints such as mortality, reproduction, or growth. These RQ values
are all below 1 for all species and corresponding trophic levels represented (Table 4-17). The earthworm
and American kestrel are important tools in this screening-level trophic transfer analysis as they
represent an animal with direct ingestion of soil (i.e., the earthworm) and as a top avian predator (i.e.,
the kestrel). Hazard values representing effects at the sub-organ level were identified for the earthworm
(alterations in gastrointestinal tract) and American kestrel (alterations in plasma thyroid hormone
levels). TCEP in biota calculated for the earthworm and American kestrel are at doses of 0.0055 and
0.0016 mg/kg/day, respectively, for the highest modeled soil TCEP concentration with a production
volume of 2,500 lb/year. They did not equal or exceed these species hazard thresholds described within
Section 4.2.4.2. The hazard value for the American kestrel (doses of 0.0025 mg/kg/day) did not result in
any detectable impacts to ecologically relevant endpoints of body weight or food consumption from this
21-day dietary exposure study with TCEP (Fernie et al.. 2015). One COU (i.e., Use in paints and
coatings at job sites) at the 25,000 lb/year production volume resulted in TCEP concentrations of 0.025
mg/kg/day; however, this production volume is believed to be an overestimate of current production
volumes in the United States (see Section 1.1.1). In addition, the screening-level analysis used equation
terms (e.g., area use factor and the proportion of TCEP absorbed from prey and soil) all set to the most
conservative values further emphasizing a cautious approach to risk to TCEP via trophic transfer.
Table 4-17. RQs for Screening-Level Trophic Transfer of Soil TCEP in Terrestrial Ecosystems
Using EPA's Wild
ife Risk Model for Eco-SSLs"
Organism
TCEP Concentration
in Biota
(mg/kg/day)6
Hazard
Threshold
(mg/kg-bw/day)
Reference for Hazard
Value or TRV
(Overall Quality
Determination)
RQ
Nematode
(Caenorhabditis
elegans)
0.0055
612
(Xu et al.. 2017) (Hiah)
9.0E-06
Mammal
0.004
44
N/Ac
9.8E-05
Woodcock
(Scolopctx minor)
0.005
N/A
N/A'#
N/A
11 Calculated using highest modeled soil TCEP concentrations with a production volume of 2,500 lb/year (0.0055
mg/kg); see also Equation 4-1.
h TCEP concentration represents the highest modeled soil concentration via AERMOD modeling with a production
volume of 2,500 lb/year.
c Mammal TCEP TRV value calculated using several studies as per (U.S. EPA. 2007a).
'' No TCEP hazard threshold value for this representative species is available.
Risk estimates were calculated for the highest releasing COUs at each production volume scenario for
TCEP soil concentrations resulting from the application of biosolids (see Section 3.3.3.5). The high-end
estimates with the 2,500 lb/year production volume scenario for Processing/ Incorporation into
formulation, mixture, or reaction product/ Paint and coating manufacturing COU resulted in trophic
transfer screening-level RQ values below 1 for both the soil invertebrate and mammal (Table 4-18). The
same COU with central tendency estimates from the 25,000 lb/year production volume scenario also
resulted in RQ values below 1 for the soil invertebrate and mammal (Table 4-18).
Page 141 of 638
-------
Table 4-18. RQs for Screening-Level Trophic Transfer of Biosolid TCEP in Terrestrial
Ecosystems Using EPA's Wildlife Risk Model for Eco-SSLs"
Organism
TCEP Concentration
in Biota
(mg/kg/day)6
Hazard
Threshold
(mg/kg-bw/day)
Reference for Hazard
Value or TRV
(Overall Quality
Determination)
RQ
Processing/Incorporation into formulation, mixture, or reaction product/ Paint and coating manufacturing COU,
high-end estimate from 2,500 lb/year PV scenario
Nematode
(Caenorhabditis
elegans)
0.141
612
(Xu et al.. 2017) (Hiah)
2.3E-04
Mammal
0.112
44
N/Ac
2.5E-03
Woodcock
(Scolopax minor)
0.140
N/A
N/A'#
N/A
Processing/Incorporation into formulation, mixture, or reaction product/ Paint and coating manufacturing COU,
central tendency estimate from 25,000 lb/year PV scenario
Nematode
(Caenorhabditis
elegans)
0.529
612
(Xu et al.. 2017) (Hiah)
8.6E-04
Mammal
0.420
44
N/Ac
9.5E-03
Woodcock
(Scolopax minor)
0.527
N/A
N/A'#
N/A
a Calculated using highest modeled biosolids TCEP concentrations with a production volumes of 2,500 lb/year
(0.141 mg/kg) and 25,000 lb/year (0.529 mg/kg); see also Equation 4-1.
h TCEP concentration represents the highest modeled biosolid concentration as per Section 3.3.3.5.
c Mammal TCEP TRV value calculated using several studies as ocr (U.S. EPA. 2007a).
'' No TCEP hazard threshold value for this representative species is available.
There are no reported studies within the pool of reasonably available published literature that quantify
TCEP soil concentrations in the United States. A study with an overall quality determination of high
monitored TCEP soil concentrations in the summer (August) and winter (January and February) months
in Germany (Mihailovic and Fries. 2012). The soil collection site was characterized as being located
approximately 3 km from the city center of Osnabrueck and about 20 m from buildings constructed of
reinforced concrete with facades predominately comprised of glass. Biota concentrations and RQ values
were calculated using the same assumptions as described previously in Table 4-10, utilizing the highest
TCEP soil concentration reported in Mihailovic and Fries (2012). Note that this study should be
considered to represent TCEP concentrations in soil from an ambient urban environment and is not
directly comparable to scenarios detailed within the current risk evaluation. In a related study at the
same site, the authors postulated that TCEP concentrations resulted from atmospheric deposition and
potentially from cars, and emphasizing the importance of considering atmospheric deposition of
chlorinated organophosphate esters (e.g., TCEP) in future risk assessments (Mihailovic et al.. 2011). The
RQs are below 1 for all species and corresponding trophic level represented (Table 4-19). TCEP
concentrations in biota calculated for the earthworm and American kestrel were 5.89x 10 3 and
1.70xl0~3 mg/kg/day, respectively, and do not equal or exceed these species hazard thresholds described
in Section 4.2.4.2.
Page 142 of 638
-------
Table 4-19. RQs Calculated with Highest Mean TCEP Soil Concentration (5.89E-03 mg/kg) from
Monitored Values in Published Literature for Screening-Level Trophic Transfer of TCEP in
Terrestrial Ecosystems Using EPA's Wildlife Risk Model for Eco-SSLs"
Organism
TCEP
Concentration in
Biota (mg/kg/day)6
Hazard
Threshold
(mg/kg-bw/day)
Reference for Hazard
Value or TRV
(Overall Quality
Determination)
RQ
Nematode
(Caenorhabditis elegans)
5.89E-03
612
Xu et al. (2017) (Hieh)
9.6E-06
Mammal
4.60E-03
44
N/Ac
1.0E-04
11 As reported in (Mihailovic and Fries. 2012); see also Eauation 4-1
h TCEP concentration represents the highest mean recorded soil concentration (5.89E-03 mg/kg) as reported in
(Mihailovic and Fries. 2012)
c Mammal TCEP TRV value calculated using several studies as detailed in (U.S. EPA. 2007a)
J No TCEP hazard threshold value for this representative species is available
RQs were below 1 for semi-aquatic terrestrial receptors via trophic transfer from fish and the resulting in
the highest modeled TCEP surface water concentrations (Processing/ Incorporation into article/
Aerospace equipment and products and automotive articles and replacement parts containing TCEP,
Table 4-20). RQ and biota concentration values for all COUs are presented within Table Apx H-14. The
hazard confidence for the chronic mammalian assessment was reported as robust within Section 4.2.6
and BCF values used to approximate TCEP concentrations within fish were from a high-quality study
(Arukwe et al.. 2018). The modeled TCEP concentrations within this analysis are five orders of
magnitude greater than surface water concentrations identified from the WQP database and the
published literature (Table 4-13 and Table 4-14). These results align with previous risk assessments that
concluded that TCEP is not viewed as a bioaccumulative compound (U.S. EPA. 2015a; EC. 2009; ECB.
2009).
Table 4-20. Selected RQs (Highest Fish TCEP Concentrations) Based on Potential Trophic
Transfer of TCEP from Fish to American Mink (Mustela vison) as a Model Aquatic Predator
Using EPA's Wildlife Risk Model for Eco-SSLs"
COU
Production
Volume
(lb/year)
Release
Distribution
SWCfl
(ppb)
Fish
Concentration
(mg/kg)
American Mink
(Mustela vison)
TCEP in
Biota
(mg/kg/day)
RQ
Processing/Incorporation
into article/Aerospace
equipment and products
and automotive articles
and replacement parts
containing TCEP
2,500
High-End
10,900
3.71
2.34
0.08
11 See also Equation 4-1
h TCEP Surface Water Concentration (SWC) calculated using VVWM-PSC
Page 143 of 638
-------
4,3,5 Connections and Relevant Pathways from Exposure Media to Receptors
4.3.5.1 Aquatic Receptors
Surface Water, Benthic Porewater, and Sediment
Within the aquatic environment, a two-tiered modeling approach was employed to predict surface water,
pore water, and sediment TCEP concentrations. If E-FAST 2014 predicted 7Q10 surface water
concentrations were greater than the chronic or acute COCs, the VVWM-PSC model was then used to
confirm whether the predicted surface water concentration days of exceedance as determined by the
acute COC and chronic COC. For TCEP, all five applicable COUs (Table 4-9) modeled in E-FAST
produced chronic RQ values greater or equal to 1, prompting the use of VVWM-PSC for greater
ecological resolution on TCEP concentrations and days of exceedance within the water column and
benthic compartments (see Section 4.3.1).
Air Deposition to Water and Sediment
EPA used IIOAC and AERMOD to estimate air deposition from hypothetical facility releases and to
calculate pond water and sediment concentrations 1,000 m from the hypothetical facility. Pond water
concentrations from air deposition were estimated for the COUs with air releases (Table 4-9). The
highest estimated 95th percentile pond water concentration from annual deposition, across all exposure
scenarios, was 0.49 ppb for the Commercial use of paints and coatings scenario at an annual production
volume of 2,500 lb per year. This highest modeled concentration within a pond at 1,000 m from a point
source was which is 4 orders of magnitude less than the lowest surface water concentration modeled
using the model, VVWM-PSC (2,350 ppb as a maximum 1-day average concentration for the
Manufacture/Import/Import COU at an annual production volume of 2,500 lb per year). Air deposition
to sediment as reported in Section 3.3.2.10 indicated the highest annual deposition at 1,000 m was 125
ppb, which is about seven times lower than the lowest sediment TCEP value modeled with VVWM-PSC
(Incorporation into paints and coatings - solvent borne at 893 ppb) and about 40 times lower than the
highest PSC value for laboratory chemicals (5,040 ppb). Using VVWM-PSC, sediment concentrations
from aquatic releases of TCEP ranged from 893 ppb to 5,040 ppb for the production volume of 2,500
lb/year, respectively, and represent a significant driver of TCEP deposition to sediment within flowing
water systems. Although the IIO AC and AERMOD were applied to a generic farm pond setting to
calculate concentrations of TCEP in pond surface water and pond sediment, these models do not account
for media exchange of the chemical of interest as is the case for VVWM-PSC. It is not anticipated that
air deposition to water significantly contributes to TCEP concentrations within flowing receiving waters.
TCEP Runoff from Biosolids
Due to its persistence, it is likely that dissolved TCEP will eventually reach surface water via runoff
after the land application of biosolids. A review of reasonably available literature indicates that modeled
surface water, pore water, and sediment concentrations are approximately half the highest concentrations
and approximately 50 times greater than the mean values biosolid concentrations reported in Wang et al.
(2019c). Direct exposure of TCEP to aquatic receptors via biosolids was not assessed quantitatively (see
Section 3.3.3).
4.3.5.2 Terrestrial Receptors
Dermal Contact and Inhalation by Wildlife
Direct exposure of TCEP to terrestrial receptors via air was not assessed quantitatively because dietary
exposure was determined to be the driver of exposure to wildlife. The contribution of exposure risk from
inhalation relative to the ingestion exposure route is not expected to drive risk because of dilution
associated environmental conditions and the deposition of TCEP from air to soil (U.S. EPA 2003a. b).
AERMOD results indicate a maximum ambient air concentration (95th percentile, MetHIGH) of
Page 144 of 638
-------
6.08x 1CT7 [j,g/m3 at 1,000 m from a hypothetical facility for the Use of paints and coatings - spray
application OES under the 2,500 lb/year production volume using the Suburban forest land category
scenario (see Section 3.3.1.2). AERMOD results for the same conditions and COU for air deposition to
soil indicate a TCEP concentration of 5.58 (J,g/kg at 1,000 m from a hypothetical facility (Table Apx
H-12). In addition, TCEP is not persistent in air due to short half-life in the atmosphere (ti/2 = 5.8 hours)
(U.S. EPA. 2017a) and because particle-bound TCEP is primarily removed from the atmosphere by wet
or dry deposition (see Section 4.1.3.2).
Based on the Guidance for Developing Ecological Soil Screening Levels (U.S. EPA. 2003 a. b)
document, for terrestrial wildlife, relative exposures associated with inhalation and dermal exposure
pathways are lower, even for volatile substances, compared to direct ingestion and ingestion of food (by
approximately 1,000-fold). In addition, TCEP is not expected to bioaccumulate in tissues, and the
screening-level trophic transfer analysis indicated that concentrations will not increase from prey to
predator in either aquatic or terrestrial food webs (see Appendix F.2.6).
Biosolids
TCEP is released to the environment by various exposure pathways (Figure 2-1). The exposure pathway
for terrestrial organisms is through soil. Deposition of TCEP from air to soil is the primary exposure
pathway. A secondary source of TCEP contamination in soil is from the application of biosolids. Risk
estimates were calculated for the highest releasing COUs at each production volume scenario for TCEP
soil concentrations resulting from the application of biosolids (see Section 3.3.3.5) within a screening-
level trophic transfer analysis. Results from screening the highest COUs at both the 2,500 lb/year and
25,000 lb/year production volume scenarios resulted in RQs below one for all organisms and both PV
scenarios (see Section 4.1.4).
Air Deposition to Soil
As described in Section 3.3.3.2, IIOAC and subsequently AERMOD were used to assess the estimated
release of TCEP via air deposition from specific exposure scenarios to soil (Table 4-9). Estimated
concentrations of TCEP that could be deposited in soil via air deposition at the community level (1,000
m from the source) exposure scenarios have been calculated (see Section 4.3.1).
Soil in Diet
Following the basic equations as reported within Chapter 4 of EPA's Guidance for Developing
Ecological Soil Screening Levels, wildlife receptors may be exposed to contaminants in soil by two main
pathways: incidental ingestion of soil while feeding, and ingestion of food items that have become
contaminated due to uptake from soil (U.S. EPA. 2005a). Within this model, incidental oral soil
exposure is added to the dietary exposure resulting in total oral exposure greater than 100 percent (see
Section 4.1.4).
Surface Water Ingestion in Wildlife
Because surface water sources for wildlife water ingestion are typically ephemeral, the trophic transfer
analysis for terrestrial organisms assumed TCEP exposure concentration for wildlife water intake are
equal to soil concentrations for each corresponding exposure scenario (see Section 4.1.4).
For semi-aquatic terrestrial species, the TRV was used with the American mink for the screening-level
assessment (Table 4-10). Similar to the soil concentrations used as term Soih in Equation 4-3, the
highest surface water concentration modeled via VVWM-PSC was used as a surrogate for the TCEP
concentration found in the American mink's diet (see Section 4.3.1.1).
Page 145 of 638
-------
Semi-aquatic Wildlife
The American mink was used as the representative species for semi-aquatic mammals. As a
conservative assumption, 100 percent of the American mink's diet is predicted to come from fish. Fish
concentration (mg/kg) was calculated using surface water concentrations of TCEP from VVWM-PSC
assuming a BCF of 0.34 as reported for whole body values from 1 mg/L TCEP exposures under
laboratory conditions (Arukwe et al.. 2018). The conservative approach for calculated fish tissue
concentrations presented in Section 4.1.2.2 was utilized for trophic transfer analysis to semi-aquatic
mammals (see Section 4.3.1.1).
4.3.6 Summary of Environmental Risk Characterization
4.3.6.1 COUs/OESs with Quantitative Risk Estimates
EPA had uncertainty in the production volume; however, even at the realistic production volume of
2,500 lb/year, EPA found chronic RQs above 1 with more than 30 days of exceedance for aquatic
receptors from both TCEP concentrations in surface water and pore water modeled with VVWM-PSC.
Additionally, because of the physical and chemical and fate properties, EPA expects TCEP to partition
between water and sediment and be persistent within the sediment compartment. Therefore, EPA has
moderate confidence that there is risk to aquatic organisms for 5 out of 20 COUs.
The current environmental risk characterization on TCEP utilizes two alternate production volume
assumptions for the calculation of RQ values. The 25,000 lb/year production volume is used as the high-
end estimation. It is based on the reporting threshold for TCEP in CDR; however, given EPA's research,
this is believed to be an overestimate of current production volumes in the United States. The 2,500 lb
production volume is reflective of estimated current production volumes. In the current section, the
analyses using 2,500 lb/year production volume are presented. Table 4-21 and Table 4-22 present RQ
values for exposure scenarios with a production volume of 2,500 lb/year and corresponding
environmental risk for aquatic and terrestrial receptors, respectively. Exposure data and corresponding
RQ values produced with a production volume of 25,000 lb/year are presented within the Appendix H.
Exposure concentrations were modeled based on COU related releases to the aquatic environment and
are represented by TCEP values within surface water and pore water. Confidence in aquatic exposure
estimates is "moderate" with modeling parameters considering inputs from COUs and physical and
chemical and fate parameters specific to TCEP. Surface water monitoring data were available from the
WQP database and published literature. The overall exposure confidence for acute and chronic aquatic
assessment were both rated as "moderate" (Table 4-24) with the inclusion of physical and chemical
parameters represented within models performed with VVWM-PSC. The VVWM-PSC model identified
substantial deposition of TCEP to the benthic compartment, which comprises sediment and benthic pore
water. Physical and chemical properties including but not limited to Koc, benthic half-life, and
hydrolysis half-life within the VVWM-PSC model, aligns with the partitioning to organic carbon in
sediment (see Appendix F.2.3.2) and persistence (see Appendix F.2.3.1). These parameters resulted in
modeled data indicating TCEP concentrations residing within pore water over longer durations of time
(days of exceedance) when compared to results from surface water concentrations for the chronic COC
(2.8 ppb).
Within the aquatic environment, chronic RQs for aquatic receptors from TCEP exposure are elevated
above one and have corresponding days of exceedance greater than 30 days within both surface water
and pore water. Chronic RQ values for COUs with releases to water are presented with several scenarios
for flow and production volume. Table 4-21 demonstrates RQ values from high-end production
estimates from a 2,500 lb/year scenario with a 7Q10 low flow release representing the 50th percentile
Page 146 of 638
-------
stream flows of all facilities in a given industry sector, as defined by the SIC codes of the industry
sector. Tables within Appendix H represent modeled surface water (TableApx H-2) and pore water
(TableApx H-3) using central tendency production estimates from the 2,500 lb/year scenario with the
same 50th percentile 7Q10 flows. Both central tendency and high-end estimate scenarios with the 2,500
lb/year production volume were found to have chronic RQs greater than 1 with days of exceedance
greater than 30 for 4 COUs with quantified releases to water. The RQ value for the central tendency
estimate for the paints and coatings COU with a P50 7Q10 low flow condition resulted in an RQ value
of 6.7; however, the days of exceedance were 29. The COC for this RQ calculation was derived from a
TCEP exposure of 30 days, thus the COC for this COU was 1 day short of exceeding that
duration. Acute RQs for these production volume scenarios and under 1 for all five COUs with
quantified releases to water.
When modeling TCEP concentrations in surface water and pore water with a 7Q10 low flow release
representing the 90th percentile stream flows of all facilities in each industry sector, both the high-end
and central tendency estimates resulted in chronic RQs less than 1 and days of exceedance less than 30
for all five COUs with quantified releases to water. Results for these scenarios with the 90th percentile
7Q10 stream flows and high-end estimates are presented in Table Apx H-6 and Table Apx H-7 for
surface water and pore water, respectively. Results of the 90th percentile 7Q10 stream flows and central
tendency estimates are presented in Table Apx H-4 and Table Apx H-5 for surface water and pore
water, respectively. These results indicate the critical role of receiving water flow as an input in
determining TCEP concentrations within both surface water and pore water.
For pore water, chronic RQs were greater than or equal to 1 with over 30 days of exceedance for all five
relevant COUs (Table 4-21). Days of exceedance were greater in pore water (Table 4-12) than surface
water (Table 4-11), indicating that TCEP is a more persistent hazard to benthic dwelling organisms with
increased durations of exposure. All relevant COCs and relevant flow data for VVWM-PSC results for
modeled pore water concentrations are available in Table 4-12. Concern for these RQs within pore water
is the lasting effects on benthic biota and potential community-level impacts from chronic TCEP
exposure within this aquatic compartment. Many benthic invertebrates are detritivores, meaning they
feed on dead plant and animal material or contribute to the liberation of additional nutrient resources by
further breaking down these materials. These detritivorous benthic invertebrates often serve as an
important food source for many juvenile fishery and non-game resident species.
Chronic RQs were greater than one with over 30 days of exceedance for surface water and pore water
TCEP modeled via VVWM-PSC at the 2,500 lb/year production volume for all five relevant COUs (Life
Cycle Stage/Category/Subcategory/OES):
• Manufacturer/Import/Import/Repackaging;
• Processing/Incorporation into formulation, mixture, or reaction product/Paint and coating
manufacturing/Incorporation into paints and coatings - 1-part coatings and 2-part reactive
coatings
• Commercial use/Paints and coatings/Paints and coatings/Use in paints and coatings - Spray
application
• Processing/Incorporation into formulation, mixture, or reaction product/Polymers used in
aerospace equipment and products/Processing into 2-part resin article; and
• Commercial use/Laboratory chemicals/Laboratory chemicals/Lab chemical - Use of laboratory
chemicals.
Page 147 of 638
-------
Table 4-21. Exposure Scenarios (Production Volume of 2,500 lb TCEP/year) and Corresponding Environmental Risk for Aquatic
Receptors with TCEP in Surface Water and Pore Water
cou
OESai
Aquatic Receptors'
Surface Water
Pore Water
Life Cycle Stage/Category
Subcategory
Acute
RQrf
Conf in
Acute RQ
Inputs^
Chroni
c
RQf
DoE®
Conf in
Chronic
RQ
Inputs^
Acute
RQrf
Conf in
Acute
RQ
Inputs^
Chroni
cRQf
DoE®
Conf in
Chronic RQ
Inputs^
Manufacture/Import
Import
Repackaging
0.14
Moderate
112.50
34
Moderate
9.16E-03
Moderate
44.28
252
Moderate
Processing/Incorporation into
formulation, mixture, or
reaction product
Paint and coating
manufacturing
Incorporation into
paints and coatings
- 1-part coatings
0.60
Moderate
244.29
80
Moderate
0.02
Moderate
96.78
298
Moderate
Processing/Incorporation into
formulation, mixture, or
reaction product
Paint and coating
manufacturing
Incorporation into
paints and coatings
- 2-part reactive
coatings
0.49
Moderate
110.71
33
Moderate
9.10E-03
Moderate
43.92
251
Moderate
Processing/Incorporation into
formulation, mixture, or
reaction product
Polymers used in
aerospace
equipment and
products
Formulation of
TCEP into 2-part
reactive resins
0.33
Moderate
132.50
33
Moderate
1.09E-02
Moderate
52.50
260
Moderate
Commercial use/Paints and
coatings
Paints and
coatings
Use in paints and
coatings - Spray
application
0.54
Moderate
129.28
52
Moderate
1.06E-02
Moderate
51.40
262
Moderate
Commercial use/Laboratory
chemicals
Laboratory
chemicals
Lab chemical -
Use of laboratory
chemicals
5.8E-02
Moderate
34.46
210
Moderate
5.42E-03
Moderate
32
364
Moderate
Modeled TCEP concentrations and RQ values for all relevant exposure scenarios are available in Table 4-11, and Table 4-12
" Production volume of 2,500 lb TCEP/yr uses high-end estimates (95th percentile for all COUs except the laboratory chemicals COU uses the 1st percentile)
h Risk assessed to aquatic receptors based on TCEP releases from wastewater, WQP database, and published literature
c All exposure values and Days of Exceedance (DoE) modeled using WWM-PSC
d Acute RQ derived using a Concentration of Concern of 16,700 ppb
Conf = Confidence. Confidence in Acute RQ or Chronic RQ inputs is detailed in Section 4.3.7.2
' Chronic RQ derived using a Primary Concentration of Concern of 2.8 ppb
g Days of Exceedance (DoE) modeled using WWM-PSC
Page 148 of 638
-------
Table 4-22. Exposure Scenarios (Production Volume of 2,500 lb TCEP/year) and Corresponding Environmental Risk for Terrestrial
Receptors with TCEP in Soil (Invertebrates) and Trophic Transfer
cou
OES"
Meteorological
Model*
Terrestrial Receptors'
Life Cycle Stage/Category
Subcategory
Soil (Invertebrates)''
Trophic Transfer (Soil)''
Trophic Transfer
(Water)'
RQ
Conf. in
RQ Inputs^
Mammal RQ
Conf. in RQ
Inputs^
American
Mink RQ
Conf. in
RQ Inputs^
Manufacture/Import
Import
Repackaging
MetCT
2.4E-06
Moderate
1.8E-06
Robust
0.02
Robust
MetHI
3.1E-09
2.3E-06
Processing/Incorporation
into formulation, mixture, or
reaction product
Paint and coating
manufacturing
Incorporation into
paints and coatings -
1-part coatings
MetCT
5.4E-08
Moderate
4.0E-05
Robust
0.08
Robust
MetHI
9.3E-08
6.8E-05
Processing/Incorporation
into formulation, mixture, or
reaction product
Paint and coating
manufacturing
Incorporation into
paints and coatings -
2-part reactive
coatings
MetCT
1.8E-08
Moderate
1.3E-05
Robust
0.07
Robust
MetHI
3.9E-08
2.9E-05
Processing/Incorporation
into formulation, mixture, or
reaction product
Polymers used in
aerospace equipment
and products
Formulation of TCEP
into 2-part reactive
resins
MetCT
2.0E-08
Moderate
4.7E-05
Robust
0.08
Robust
MetHI
4.2E-08
4.6E-05
Processing/Incorporation
into article
Aerospace equipment
and products and
automotive articles
and replacement parts
containing TCEP
Processing into 2-part
resin article
MetCT
6.4E-08
Moderate
1.5E-05
Robust
NA
Robust
MetHI
6.3E-08
3.1E-05
Commercial Use/Paints and
coatings
Paints and coatings
Use in paints and
coatings at job sites
MetCT
6.5E-06
Moderate
0.005
Robust
0.04
Robust
MetHI
9.1E-06
0.007
Commercial Use/Laboratory
chemicals
Laboratory chemicals
Lab chemical - Use of
laboratory chemicals
MetCT
7.9E-08
Moderate
5.8E-05
Robust
7.0E-04
Robust
MetHI
7.6E-08
5.6E-05
" Production volume of 2,500 lb TCEP/year uses high-end estimates (95th percentile for all COUs except the laboratory chemicals COU uses the 1st percentile)
b The ambient air modeled concentrations and deposition values are presented for two meteorology conditions (MetCT: Sioux Falls, South Dakota, for central tendency
meteorology; and MetHI: Lake Charles, Louisiana, for higher-end meteorology)
c Risk assessed to terrestrial receptors based on TCEP releases as fugitive air and stack air deposition to soil, trophic transfer, and published literature.
J Estimated concentrations of TCEP (90th percentile) that could be in soil via air deposition at a community (1,000 m from the source) exposure scenario
e Fish concentration (mg/kg) was calculated using surface water concentrations of TCEP from WWM-PSC assuming a BCF of 0.34 as reported for whole body values from 1
me/L TCEP exposures under laboratory conditions (Arukwe et al.. 2018)
f Conf = Confidence; Confidence in RQ inputs are detailed in Section 4.3.7.2
Page 149 of 638
-------
RQs were less than 1 for all relevant COUs for air deposition to soil at 1,000 m (Table 4-22). The
highest soil concentration from AERMOD predictions is 0.0055 mg/kg based on TCEP use in Paints and
coatings at job sites at 1,000 m with the 2,500 lb/year production volume and higher-end meteorology
condition. There are no published literature or monitoring databases with TCEP soil concentrations from
U.S. sites and one comparative study from Germany (Mihailovic and Fries. 2012). RQs for soil
invertebrates were less than 1 with soil TCEP concentrations as reported for different sample periods
from Mihailovic and Fries (2012) (Table 4-16). This study should be considered to represent TCEP
concentrations in soil from an ambient urban environment and is not directly comparable to scenarios
detailed within the current risk evaluation. Mihailovic et al. (2011) emphasized the importance of
atmospheric deposition of chlorinated organophosphate esters in risk assessments, which the current risk
evaluation has taken into consideration for environmental risk characterization.
Trophic transfer of TCEP and potential risk to terrestrial animals was based on modeled soil data from
AERMOD and concentrations reported within Mihailovic and Fries (2012). A screening-level approach
was conducted as described in EPA's Guidance for Developing Ecological Soil Screening Levels (U.S.
EPA. 2005a). The two analyses performed represented: (1) trophic transfer for animals from exposures
originating with TCEP soil concentrations and terrestrial prey items (Table 4-19), and (2) trophic
transfer based for animals from exposures with TCEP water concentrations and aquatic prey items
(Table 4-20). Table 4-22 demonstrates that RQs were less than 1 for any modeled soil concentrations
and COUs based on the chronic hazard threshold for terrestrial invertebrate identified in Appendix H.
The chronic TRV, calculated using empirical toxicity data with mice and rats, also demonstrated RQs
less than 1 for all modeled soil concentrations (Table 4-22). In addition, RQs were less than 1 for all
species represented within trophic levels using TCEP soil concentrations reported within Mihailovic and
Fries (2012) (Table 4-19). For semi-aquatic animals, RQs were also less than 1 for semi-aquatic
terrestrial mammals via trophic transfer from fish and the highest modeled TCEP surface water
concentrations (Table 4-20). The results of these screening-level trophic transfer analyses corroborate
previous risk assessments indicating TCEP is not a bioaccumulative compound (U.S. EPA. 2015a; EC.
2009; ECB. 2009).
In the current environmental risk characterization for aquatic and terrestrial organisms, EPA considered
aggregating exposure that a population would experience from being in close proximity to multiple
facilities releasing TCEP to the environment. However, EPA did not aggregate across facilities for
environmental exposures or risk because location information was not reasonably available for facilities
releasing TCEP to the environment. Environmental media concentrations from monitoring data (i.e., not
associated with a specific exposure scenario or COU) were not aggregated with modeled environmental
media concentrations associated with a specific exposure scenario or COU. TCEP from monitored
surface water data reported within the WQP indicated a mean of 0.33 ± 0.02 ppb (see Section 4.3.2).
Table 4-13 demonstrates that this mean surface water concentration for TCEP resulted in acute and
chronic RQ values of 1.05x10 5 and 0.11, respectively. Similar database monitoring information were
not reasonably available for sediment TCEP concentrations; however, the model used to predict surface
water, sediment, and porewater TCEP concentrations was inclusive of physical and chemical properties
(i.e., Kow, Koc, water column half-life, photolysis half-life, hydrolysis half-life, and benthic half-life)
known to contribute to TCEP's persistence within these media.
EPA also considered aggregating across pathways of exposure for aquatic and terrestrial organisms, but
did not, because releases of TCEP to surface water and sediment were found to significantly contribute
to these media when compared to deposition to water and/or sediment via air (see Section 4.3.5.1).
Similarly, the most significant pathway for exposure to terrestrial receptors is via soil, which was
modeled from air deposition (see Section 4.3.5.2). For aquatic organisms, surface water and sediment
Page 150 of 638
-------
pathways involve primary exposure routes such as epithelial uptake (skin, gills) and oral uptake.
Aggregation of exposures via both surface water and dietary exposure was not conducted for aquatic
organisms because TCEP is not expected to bioaccumulate except at very high concentrations that could
result in risk directly from surface water (see Appendix F.2.6). The screening-level trophic transfer
analysis performed included TCEP within prey in addition to soil ingestion for terrestrial receptors and
water ingestion for semi-aquatic mammals (see Section 4.3.1.1).
4.3.6.2 COUs/OESs without Quantitative Risk Estimates
The following section represents a qualitative discussion of those remaining COUs and subsequent
OESs lacking quantitative risk estimates.
Recycling and Distribution in Commerce
EPA did not have sufficient data to estimate releases to the environment for the following COUs:
• Processing - Recycling
• Distribution in commerce
EPA was not able to quantify releases of TCEP to the environment during the recycling of e-waste. E-
waste recycling activities include receiving e-waste at the facility, dismantling or shredding the e-waste,
and sorting the recycled articles and generated scrap materials (NIOSH. 2018; Yang et al.. 2013; Siodin
et al.. 2001). Only a subset of e-waste recycling facilities is expected to handle TCEP-containing
products. The exact number of these facilities is unknown, and data were not reasonably available on the
volume or source of TCEP contained in electronics processed at any of the facilities identified.
EPA did not find reasonably available data to quantify environmental releases of TCEP from e-waste
recycling facilities. The total releases are expected to be low because TCEP is not typically used in
electronics (Stapleton et al.. 2011). Multiple studies show detections of TCEP at electronics and
electrical equipment waste (e-waste) recycling facilities at air concentrations ranging from l.OxlO"7 to
l.lxlO"3 mg/m3, though the source of the TCEP at each facility is not specified (NCBI. 2020; Grimes et
al.. 2019; S tubbings et al.. 2019; NIOSH. 2018; Yang et al.. 2013; Siodin et al.. 2001). The low air
concentration within facilities helps provide insight as to why one electronic recycling company
categorized TCEP as "less commonly used in electronics now and in the past" with a detection
percentage of 18 percent and range of "not detectable" to 10 ng/m3 resulting in TCEP not being
quantified in the NIOSH Report on Metals and Flame Retardants (Grimes et al.. 2019). The
concentrations at the site were based on personal air sampling for 19 participants over 2 days (Grimes et
al.. 2019). TCEP-containing materials from the recycling process are typically treated or disposed
following the initial processing and not reprocessed or reused (Yang et al.. 2013). EPA did not identify
any reasonably available data for the weight fraction of TCEP in e-waste. This qualitative analysis
indicates that releases of TCEP to the environment are potentially present from the recycling of e-waste.
However, under similar exposure durations and exposure frequencies reported in the current section,
releases from e-waste recycling facilities are expected to be lower relative to other quantified scenarios
corresponding to environmental risk for terrestrial receptors (see Table 4-22) with the recycling COU
expected to have lower risk than the quantified scenarios described within Section 4.3.6.1.
For purposes of assessment in this risk evaluation, distribution in commerce of TCEP consists of the
transportation associated with the moving of sealed containers of TCEP from import sites to downstream
processing and use sites, or for final disposal of TCEP. The steps of loading and unloading that are
assessed during other COUs/OESs consists of unloading TCEP into the formulation process and loading
refers to packaging the finished product prior to shipment. Loading and unloading activities that occur
during a distribution in commerce scenario would only refer to loading or unloading sealed containers
Page 151 of 638
-------
from a transport vehicle. EPA expects, under standard operating procedures, that environmental releases
from sealed containers are not expected to occur.
Aerospace Equipment and Products
EPA does not expect significant releases to the environment for the following COUs/OESs:
• Industrial use - Other use - Aerospace equipment and products and automotive articles and
replacement parts containing TCEP
o OES: Installing article (containing 2-part resin) for aerospace applications (electronic
potting)
• Commercial use - Other use - Aerospace equipment and products
o OES: Installing article (containing 2-part resin) for aerospace applications
Specifically, EPA does not expect significant releases to occur during the installation of TCEP -
containing aircraft and aerospace articles into or onto the relevant transportation equipment. The release
assessment for these COUs/OESs are provided within Section 3.6.3 of the Engineering Supplemental
file and summarized here. After TCEP-containing resins have cured, EPA expects TCEP release will be
limited by the hardened polymer matrix. Releases may occur via the mechanism of "blooming" or
volatilization from the cured resin surface during the service life of the aircraft or aerospace article, but
EPA expects that releases via this mechanism during installation activities will be negligible (OECD.
2009; NICNAS. 2001). Furthermore, installation of aerospace equipment and products would be
installed without any type of further processing of the article that would lead to potential releases
(sanding, drilling, etc.). The Agency was not able to quantify environmental releases from blooming due
to a lack of reasonably available information on the end use and service life of the product. Based on the
finding of limited environmental releases from installation of TCEP-containing aircraft and aerospace
articles into or onto the relevant transportation equipment, EPA determined that risk to the environment
is not expected from releases of TCEP during the installation of these articles.
Commercial Uses (COUs) that TCEP is no longer actively incorporated into
The COUs listed below are only linked to end of service life disposal as manufacturing and processing is
not ongoing:
• Commercial use - Furnishing, cleaning, treatment/care products - Fabric and textile products;
• Commercial use - Furnishing, cleaning, treatment/care products - Foam seating and bedding
products;
• Commercial use - Construction, paint, electrical, and metal products - Building/construction
materials - Insulation; and
• Commercial use - Construction, paint, electrical, and metal products - Building/construction
materials - Wood and engineered wood products - Wood resin composites.
EPA has confirmed from literature sources that TCEP was used for these purposes in past decades.
However, these commercial uses were phased out beginning in the late 1980s or early 1990s and
replaced by other flame retardants or flame-retardant formulations. EPA was unable to locate data to
estimate the TCEP throughput used for these products, the amounts of these products that have already
reached the end of their service life or amounts that have already been disposed of. The Agency assumes
that products with TCEP that are still in use represents a fraction of the overall amount of TCEP
previously used for these purposes and these types of products (e.g., insulation and furniture) will result
in a final deposition to landfills for disposal. TCEP releases to the environment from these commercial
use COUs are expected to be lower relative to other quantified scenarios, as these commercial COUs
have limited environmental release potential past end of service life disposal when compared to the
quantified scenarios described within Section 4.3.6.1 The consumer assessment for these articles
Page 152 of 638
-------
resulted in no consumer risk for inhalation, ingestion dermal risk for adults for the COUs with moderate
confidence. Due to the lack of reasonably available information on (1) the amount of TCEP used in
these products, (2) the amounts of these products that have already reached the end of their service life,
or (3), the amount of articles that remain in commercial environments, EPA is unable to quantify
environmental releases for commercial COUs listed above.
Processing/Incorporation into Formulation, Mixture, or Reaction Product Processing/Incorporated
into Article
EPA identified the following environmental releases via waste disposal; however, the Agency was
unable to perform quantitative risk characterization of environmental releases related to waste disposal
for the following COUs:
• Processing/Incorporation into formulation, mixture, or reaction product/Paint and coating
manufacturing;
• Processing/Incorporation into formulation, mixture, or reaction product/Paint and coating
manufacturing;
• Processing/Incorporation into formulation, mixture, or reaction product/Polymers used in
aerospace equipment and products; and
• Processing/Incorporation into article/Aerospace equipment and products
EPA was able to perform quantitative risk characterization (Table 4-9) on the COUs listed above based
on environmental releases to either fugitive or stack air and/or wastewater to on-site treatment or
discharge to POTW, where applicable (Table 3-2). Waste disposal refers to either landfill or incineration
and relies on inputs provided by the ESD or GSs. The proportion of the throughput that goes to either
landfills or incinerators was not detailed within the ESD or GS. Details pertaining to the fate of disposal
to these waste streams were unknown, a qualitative analysis of the disposal COU is presented below.
Consumer Uses
Although there is the possibility of environmental releases from consumer articles containing TCEP via
off-gassing of consumer articles, down the drain release of TCEP from domestic laundry, the end-of-life
disposal and demolitions of consumer articles, EPA was unable to quantify the environmental releases
for the following COUs:
• Consumer use - Paints and coatings;
• Consumer use - Furnishing, cleaning, treatment/care products - Fabric and textile products;
• Consumer use - Furnishing, cleaning, treatment/care products - Foam seating and bedding
products;
• Consumer use - Construction, paint, electrical, and metal products - Building/construction
materials - insulation; and
• Consumer use - Construction, paint, electrical, and metal products - Building/construction
materials - Wood and engineered wood products - Wood resin composites
EPA was unable to quantify environmental exposures from consumer releases and disposal due to
limited reasonably available information on source attribution of the consumer COUs. In previous
assessments, EPA has considered down the drain analysis for consumer products for which a reasonably
foreseen direct discharge exposure scenario can be assumed (e.g., drain cleaner, lubricant, oils). TCEP
containing dust present on consumer clothing may be released to the environment via domestic laundry;
however, due to uncertainties in the source attribution of consumer COUs to dust, and the subsequent
loading of dust on to clothing, EPA did not quantify environmental exposures for this scenario.
Consumer releases to the environment are anticipated to be dispersed and less than direct TCEP releases
Page 153 of 638
-------
from COUs/OESs quantified for risk estimates for aquatic and terrestrial receptors detailed within Table
4-9.
Disposal
TCEP was among the 10 most frequently found compounds in a study that collected wastewater from
multiple sites in the Research Triangle Park area of North Carolina between 2002 and 2005 (Giorgino et
al.. 2007). The study detected TCEP in 61.9 percent of wastewater samples, with a maximum
concentration of 0.7 ppb. The maximum concentration from the USGS study (0.7 ppb) is similar to the
maximum surface water TCEP concentration reported within published literature (0.81 ppb) used to
calculate risks (see Section 4.3.2) and resulted in RQ values of less than one for both acute and chronic
COCs (Table 4-14). The researchers indicated that flame retardants were measured primarily at sites
downstream from municipal wastewater discharges and elevated concentrations were due to surface
waters collected at a site downstream from an industrial fire.
Incineration of articles containing TCEP may create localized environmental releases. Aston et al.
(1996) reported TCEP concentrations of up to 1.95 mg/kg in pine needles (Pinus ponderosa) in the
Sierra Nevada foothills in the mid-1990s (Table 4-5). The source of the TCEP is unknown; however,
authors suspected that these levels may have been due to aerial transport and deposition from nearby
point sources such as incinerators.
The demolition and removal of commercial and consumer articles may result in environmental
exposures to TCEP. Construction waste and old consumer products can be disposed of in municipal
solid waste landfills and construction and demolition landfills. Section 3.3.3.8 models the resulting
groundwater concentration that may occur from TCEP that leaches from landfills. Section 3.3.3.5
highlights suspected leaching of TCEP from nearby landfills (Norman Landfill, Himco Dump and Fort
Devens, MA) (Buszka et al.. 2009; Barnes et al.. 2004; Hutchins et al.. 1984). Groundwater from one
well in Elkhart, Indiana, near the Himco Dump reported TCEP concentrations of 0.65 ppb to 0.74 ppb
(Buszka et al.. 2009). The Himco Dump is a closed, formerly unlicensed landfill that included a 4-acre
construction debris area. EPA issued a notice in the Federal Register finalizing the deletion of part of the
Himco Dump Superfund site from the National Priorities List (NPL) and the Indiana Department of
Environmental Management (IDEM) formally concurred with EPA's proposal on January 26, 2022.
EPA proposed the site for partial deletion in March 2022. Fort Devens is a former army installation
established in 1917 and closed in 1996 and is also an EPA superfund site. Monitoring wells down-
gradient of a land application facility near Fort Devens, Massachusetts, indicated TCEP concentrations
from 0.28 ppb to 0.81 ppb (Hutchins et al.. 1984). TCEP was detected throughout the entire length of a
leachate plume near a municipal landfill (subtitle D) near Norman, Oklahoma (Barnes et al.. 2004).
TCEP concentration detected within the groundwater plume down-gradient of the Landfill in Norman,
Oklahoma, ranged from 0.22 ppb to 0.74 ppb (Barnes et al.. 2004). Leachate samples from landfill sites
in Japan detected TCEP at ranges from 4.1 x 106 to 5.4x 109 ppb with authors indicating that plastic
wastes may serve as the origin (Yasuhara. 1995).
Without a full characterization of non-hazardous landfill (e.g., Norman Landfill) conditions and
historical wastes (e.g., Himco Dump and Fort Devens) around the country, the data needed to produce
quantitative risk estimates for disposal is not reasonably available. EPA does not have data representing
municipal and managed landfills and is uncertain how often contaminant migration occurs given modern
practices of non-hazardous landfill and historical site management. Source attribution of the consumer
uses to the leaching concentration exhibited within Sections 3.3.3.7 and 3.3.3.8 are not reasonably
available; therefore, it is unknown if these concentrations are the result of consumer and/or commercial
disposal. The possibility of environmental exposure to TCEP after the release from disposal of consumer
Page 154 of 638
-------
wastes exists. The maximum TCEP concentrations recorded within groundwater at the Norman Landfill,
Himco Dump, and Ft. Devens are 0.74 ppb, 0.81 ppb, and 0.74 ppb, respectively—which are similar to
the to the maximum surface water concentrations reported within published literature (0.81 ppb) used to
calculate risks (see Section 4.3.2) resulting in RQ values less than one for both acute and chronic COCs
(Table 4-14).
For the commercial uses that have been phased out, any currently used products that contain TCEP are
expected to be disposed of in landfills but will represent just a fraction of previous amounts from when
TCEP was used more widely. Further data are lacking with which to estimate exposure and risk from
disposal or waste treatment activities for these COUs and EPA has not quantified such risks. EPA's
confidence in these exposures is indeterminate and cannot quantify risk for the disposal or waste
treatment activities for these COUs. EPA acknowledges that while some releases and exposures could
occur during the disposal of the wide variety of items that TCEP has found its way into, based on a
review of the limited information on TCEP within groundwater at landfills and wastewater runoff
presented in the section above, these are expected to be minimal and dispersed, and exposures are
expected to be negligible.
4,3.7 Overall Confidence and Remaining Uncertainties Confidence in Environmental
Risk Characterization
The overall confidence in the risk characterization combines the confidence from the environmental
exposure, hazard threshold, and trophic transfer sections. This approach aligns with the 2021 Draft
Systematic Review Protocol (U.S. EPA 2021a) and TCEP Systematic Review Protocol (U.S. EPA
2024p). The confidence from the trophic transfer section was completed in the same manner as the
confidence in hazard threshold presented in Section 4.2.6 and Appendix G.2.3.1. For trophic transfer,
EPA considers the evidence for chronic mammalian robust, the evidence for invertebrates moderate, and
the evidence for chronic avian slight (Table 4-23). Synthesis of confidence for exposure, hazard, and
trophic transfer (when applicable) resulted in the following confidence determinations for risk
characterization RQ inputs: (1) robust for chronic mammalian evidence, (2) moderate for acute and
chronic aquatic evidence, and (3) slight for chronic avian evidence (Table 4-24).
4.3.7.1 Trophic Transfer Confidence
Quality of the Database; Strength (Effect Magnitude) and Precision
Several conservative assumptions were applied across different representative organisms within trophic
groups to represent a screening-level approach. For example, modeled TCEP concentrations within
water (VVWM-PSC) and soil (via AERMOD) were applied to all COUs. TCEP concentrations obtained
from these models were specific to each COU and production volume scenarios. Examination of
potential risk from TCEP using this hazard value should be viewed as a conservative approach
employed using both AERMOD modeled data and soil concentrations within published literature
(Mihailovic and Fries. 2012).
Trophic transfer analysis utilized American woodcock and American kestrel within the soil-based
pathway to determine potential risk from TCEP. The hazard value for the raptor species is limited to a
single study observing increased thyroid hormone production with no effects on body weight or food
consumption from a 21-day feeding study (Fernie et al.. 2015). No representative hazard data were
available for the woodcock as an avian insectivore. RQ values were not calculated for the woodcock,
which served as a prey item to the kestrel, representing uptake and transfer from a soil invertebrate to
insectivore to carnivore.
Page 155 of 638
-------
Short-tailed shrew and American mink were employed as representative species using a mammalian
TRV adjusted to their respective body weights. Mammalian hazard values for trophic transfer utilized
ecologically relevant endpoints from high-quality studies originating from human health animal model
investigations. The resulting TRV (Table 4-7) derived from mammal studies was used to calculate the
hazard threshold in mg/kg-bw. Because the TRV is scaled by body weight, smaller representative
species will have greater body burden from TCEP exposure than larger species.
For soil invertebrates, two high-quality soil invertebrate studies were available. Trophic transfer analysis
used an ecologically relevant ChV from a nematode with endpoints related to reduced growth and
shortened lifespan. The earthworm hazard value was also demonstrated in this analysis, although the
earthworm did not have an ecologically relevant endpoint effect. The earthworm is still useful for
assessing trophic transfer hazards because of its direct ingestion of soil. The earthworm also serves as a
relevant prey item for all trophic levels (i.e., short-tailed shrew, woodcock, and American kestrel).
Consistency
Inputs for soil and water TCEP concentrations displayed similarities among modeled and monitored
concentrations. The highest soil concentrations modeled via AERMOD (Table 4-15) were within one
order of magnitude to the highest soil concentrations reported within published literature (Table 4-16)
(Mihailovic and Fries. 2012). Concentrations of TCEP in whole fish reported within published literature
(Guo et al.. 2017b) represent concentrations two to three orders of magnitude lower than calculated fish
TCEP concentrations (see Section 4.1.2). Any comparison to measured values reported within published
literature should be viewed conservatively as organisms with direct proximity to the source of TCEP
release and resulting surface water concentrations as calculated using VVWM-PSC.
Biological Relevance
The use of hazard values derived from singular studies for American kestrel, earthworm, and nematode
are limiting in biological relevance; however, the application of conservative assumptions at each
trophic level ensures a cautious approach to determining potential risk. For example, if the results of the
trophic transfer show that exposure from TCEP is lower than the hazard threshold for thyroid effects,
than a qualitative assertion can be made that the exposure levels from TCEP do not indicate risk. For
avian species, only a single high-quality level study was available for the American kestrel with no
hazard value for the avian insectivore within this analysis. The short-tailed shrew and American mink
were selected as appropriate representative mammals for the soil- and aquatic-based trophic transfer
analysis, respectively (U.S. EPA 1993b). Overall, the use of exposure factors (i.e., feed intake rate,
water intake rate, the proportion of soil within the diet) from a consistent resource assisted in addressing
species specific differences within the RQ equation (U.S. EPA 1993b).
Physical and Chemical Relevance
The highest modeled TCEP concentrations for water and soil were used to investigate potential risk
from trophic transfer. Assumptions within the trophic transfer equation (Equation 4-3) for this analysis
have been considered to represent conservative screening values (U.S. EPA 2005a) and those
assumptions were applied similarly for each trophic level and representative species. Applications across
representative species included assuming 100 percent TCEP bioavailability from both the soil (AF*,-) and
biota representing prey (AFy). It is likely these considerations overrepresent TCEP's ability to transfer
among trophic levels; however, it is a precaution built into the screening-level approach (U.S. EPA
2005a).
Page 156 of 638
-------
Environmental Relevance
Although several aspects of the RQ equation were conservative and represented various species, there
are still uncertainties associated with overall relevance of this model to fit all wildlife scenarios for
potential TCEP risk. The current trophic transfer analysis investigated potential risk resulting from
TCEP exposure in media such as soil and water. This analysis was extended to represent uptake from
those media to soil invertebrates and fishes as a basis of trophic transfer from these prey to other higher
trophic levels. Analysis included TCEP soil concentrations from published literature but ultimately
relied on modeled TCEP water concentrations as the monitored TCEP values from WQP are three to
five orders of magnitude less than modeled concentrations. The area use factor is the home range size
relative to the contaminated area {i.e., site/home range = AUF with the AUF within this screening-level
analysis designated as 1 for all organisms). Application of this value in the RQ equation increases the
conservative approach to trophic transfer analysis for larger animals such as mammals and birds
assuming longer residence within an exposed area or a large exposure area.
Page 157 of 638
-------
Table 4-23. TCEP Evidence Table Summarizing Overall Con
Idence Derivet
for Trophic Transfer
Types of Evidence
Quality of the
Database
Consistency
Strength and
Precision
Biological Gradient/
Dose-Response
Relevance"
Trophic Transfer
Confidence
Aquatic
Acute Aquatic Assessment
N/A
N/A
N/A
N/A
N/A
N/A
Chronic Aquatic Assessment
N/A
N/A
N/A
N/A
N/A
N/A
Aquatic plants (vascular and algae)
N/A
N/A
N/A
N/A
N/A
N/A
Terrestrial
Chronic Avian Assessment
+
++
+
N/A
+
Slight
Chronic Mammalian Assessment
+++
++
++
N/A
++
Moderate
Terrestrial invertebrates
++
++
++
N/A
++
Moderate
11 Relevance includes biological, physical/chemical, and environmental relevance.
+ + + Robust confidence suggests thorough understanding of the scientific evidence and uncertainties. The supporting weight of scientific evidence
outweighs the uncertainties to the point where it is unlikely that the uncertainties could have a significant effect on the hazard estimate.
+ + Moderate confidence suggests some understanding of the scientific evidence and uncertainties. The supporting scientific evidence weighed against
the uncertainties is reasonably adequate to characterize hazard estimates.
+ Slight confidence is assigned when the weight of scientific evidence may not be adequate to characterize the scenario, and when the assessor is
making the best scientific assessment possible in the absence of complete information. There are additional uncertainties that may need to be considered.
Page 158 of 638
-------
4.3.7.2 Risk Characterization Confidence
Environmental risk characterization evaluated confidence from environmental exposures and
environmental hazards. Hazard confidence was represented by evidence type as reported previously in
Section 4.2.6. Trophic transfer confidence was represented by evidence type as reported in the preceding
Section 4.3.7.1. Exposure confidence has been synthesized from Section 4.1.5.1 and is further detailed
in the current section. The following confidence determinations for risk characterization RQ inputs are:
(1) robust for chronic mammalian evidence, (2) moderate for acute and chronic aquatic evidence, and
(3) slight for chronic avian evidence (Table 4-24).
Surface water concentration of TCEP were modeled initially using E-FAST and further refined using
VVWM-PSC. Refined modeling with VVWM-PSC allowed estimates of TCEP pore water and sediment
concentrations in addition to providing modeled days of exceedance for each compartment. Uncertainty
associated with location-specific model inputs (e.g., flow parameters and meteorological data) is present
as no facility locations were identified for TCEP releases.
The modeled data represent estimated concentrations near hypothetical facilities that are actively
releasing TCEP to surface water, while the reported measured concentrations represent sampled ambient
water concentrations of TCEP. Differences in magnitude between modeled and measured concentrations
may be due to measured concentrations not being geographically or temporally close to known releasers
of TCEP. VVWM-PSC allowed for the application of a standard, conservative set of parameters and
adjust for physical-chemical properties of TCEP. For example, stream reach was set to represent a
waterway with a width of 8 m and depth of 2 m.
Physical and chemical properties including, but not limited to Koc, benthic half-life and hydrolysis half-
life appear to accurately represent TCEP's persistence; however, sensitivity analysis indicated that Koc
input parameters heavily influenced the role of sediment deposition to sediment. Maruva et al. (2016)
represents an ambient environmental monitoring study within the published literature that made both
surface water and sediment collections at the same sites and similar time periods within a watershed.
Surface water collected in August and October 2013 and sediment samples collected in September 2013
were taken at 6 sites downstream of urban areas along the Santa Clara River in Southern California.
TCEP sediment concentrations were consistently one order of magnitude higher than TCEP surface
water concentrations across all sample sites. Specifically, mean (± SE) TCEP concentrations for surface
water and sediment were 0.32 ± 0.04 ppb and 2.59 ± 0.75 ppb, respectively. Although a single study,
Maruva et al. (2016) illustrates how TCEP within the water column of a flowing system can sorb to
sediment to produce elevated concentrations. The WQP data and published literature on surface water
TCEP concentrations is three to four orders of magnitude lower than modeled surface water
concentrations. Confidence in the exposure components of the RQ inputs for benthic assessment is
supported as studies within published literature are one to three orders of magnitude lower than results
obtained from VVMW-PSC modeling. Confidence in exposure parameters for surface water have been
rated "moderate" as the results are modeled from directly downstream from a hypothetical facility
releasing TCEP.
Similar to aquatic exposures for TCEP, environmental exposures to soil invertebrates, mammals, and
avian species relied on modeling air deposition to soil via AERMOD with supporting information from
published literature. The AERMOD model included two meteorological conditions (Sioux Falls, South
Dakota, for central tendency meteorology; and Lake Charles, Louisiana, for higher-end meteorology) in
addition to different production volumes (2,500 and 25,000 lb/year) to characterize potential amounts of
annual TCEP deposition to soil from air. One high-quality comparative study on TCEP soil
Page 159 of 638
-------
concentrations was identified within the published literature. TCEP fish tissue concentrations within the
Great Lakes (Guo et al.. 2017b) are two to three orders of magnitude lower than the TCEP tissue
concentrations calculated using a whole organism BCF value from another high-quality study (Arukwe
et al.. 2018). Modeled soil concentrations were within one order of magnitude of a single study from
published literature (Mihailovic and Fries. 2012); however, it is important to note that similarity with a
single study is not enough to build confidence in the relevance or accuracy of modeled results.
Table 4-24. TCEP Evidence Table Summarizing Overall Confidence for Environmental
Risk Characterization
Types of Evidence
Exposure
Hazard
Trophic
Transfer
Risk
Characterization
RQ Inputs
Ac
uatic
Acute aquatic assessment
++
++
N/A
Moderate
Chronic aquatic assessment
++
++
N/A
Moderate
Terrestrial
Chronic avian assessment
++
+
+
Slight
Chronic mammalian assessment
++
+++
++
Robust
Terrestrial invertebrates
++
++
++
Moderate
+ + + Robust confidence suggests thorough understanding of the scientific evidence and uncertainties. The supporting
weight of scientific evidence outweighs the uncertainties to the point where it is unlikely that the uncertainties could
have a significant effect on the hazard estimate.
+ + Moderate confidence suggests some understanding of the scientific evidence and uncertainties. The supporting
scientific evidence weighed against the uncertainties is reasonably adequate to characterize hazard estimates.
+ Slight confidence is assigned when the weight of scientific evidence may not be adequate to characterize the
scenario, and when the assessor is making the best scientific assessment possible in the absence of complete
information. There are additional uncertainties that may need to be considered.
Page 160 of 638
-------
5 HUMAN HEALTH RISK ASSESSMENT
EPA assessed human health risks of TCEP exposure to workers and ONUs, consumers, and the general
population. Section 5.1 describes exposures to workers and ONUs via inhalation and oral routes;
workers via dermal routes; consumers via inhalation, dermal, and oral routes; and the general population
via oral, dermal, and inhalation routes. Human health hazards, including cancer and non-cancer endpoint
identification and dose-response, are described in Section 5.2. Human health risk characterization is
described in Section 5.2.9.
Updates to this section since the draft risk evaluation was released in December 2023 include the
following:
1. Updated the evidence integration/causal descriptor for developmental toxicity from likely to
suggestive;
2. Evaluated additional epidemiology studies for neurotoxicity, kidney toxicity,
immune/hematological, thyroid, lung/respiratory, body weight, developmental toxicity, and
cancer;
3. Revised analysis of the CEM Model for consumers to account for model updates and minor
changes to parameter inputs;
4. Revised analysis for consumer paints and coatings COU;
5. Added sensitivity analysis for minimum weight fraction values; and
6. Revised dermal estimates and oral ingestion estimates from soils to include soil concentrations
from BST analysis.
Page 161 of 638
-------
5.1 Human Exposures
Human Exposures (Section 5.1):
Key Points
EPA evaluated all reasonably available information for occupational, consumer, and general
population exposure to TCEP, including consideration of the potential for increased susceptibility
across PESS considerations (see also Section 5.3.3 and Appendix D). The key points are summarized
below:
• Workers and ONUs can be exposed to TCEP via inhalation by dust or vapor.
o However, large amounts of dust are not expected to be generated based on the types of
activities that occur during the processing or use of TCEP-containing products or articles,
o Workers can also be exposed to mists generated during the spray application of TCEP-
containing paint products, but ONUs are not expected to be present during this use.
o Workers will be exposed to TCEP via dermal exposure when processing liquid TCEP.
However, once TCEP has been incorporated into an article the ability for appreciable
amounts of TCEP to be absorbed through the skin will decrease significantly as there is
little need for further processing of an article during installation.
• Chronic TCEP exposures from consumer articles to infants and children are the most relevant
duration and populations of interest. Children's mouthing activity is an important factor when
estimating exposure to TCEP in consumer products.
o For consumer exposures, the inhalation route dominates exposure for building and
construction materials such as roofing insulation, acoustic ceilings, and wood flooring.
Exposures to infants and children for fabric and textiles, foam seating and bedding
products, and wooden TV stands is dominated by the oral route,
o Inhalation exposures are highest for building and construction products due to emission
of vapors from consumer articles,
o Dermal exposures are highest for wood resin products to children,
o Ingestion exposures are highest for foam seating and bedding products for children.
• Fish ingestion is the most important exposure scenario for TCEP exposure to the general
population. The bioaccumulation factor (BAF) and fish ingestion rate are sensitive
parameters that influence these exposure estimates. Tribal populations for whom fish is
important dietarily and culturally may have higher exposures than the general population and
subsistence fishers.
• Fenceline communities may have elevated exposures from facilities that release TCEP. No
site-specific information was available for TCEP, so EPA varied several inputs to show a
range of possible exposures from a hypothetical facility.
• EPA identified several PESS groups: Infant exposure to TCEP via human milk was estimated
by considering a maternal dose due to occupational, consumer, and general population
exposures. Firefighters were identified as a PESS group through occupational exposure (see
Section 5.3.3). Children and infants were identified as PESS through consumer exposure.
Tribal communities, subsistence fishers, children, infants, and people living in fenceline
communities near facilities that emit TCEP were identified as PESS through general
population exposures.
Page 162 of 638
-------
5.1.1 Occupational Exposures
TCEP - Occupational Exposures (Section 5.1.1):
Key Points
EPA evaluated the reasonably available information for occupational exposures. The key points of
the occupational exposure assessment are summarized below:
• Occupational exposure data available for TCEP:
o EPA only identified monitoring data for dust occurring within an e-waste recycling
facility; monitoring data for the remaining COUs/OESs were not found, most likely
because TCEP does not have an assigned OSHA PEL (permissible exposure limit) and is
therefore not typically tested for in the workplace,
o For OESs that do not have data, EPA used relevant generic scenario and/or emission
scenario documents to identify worker activities and exposure routes that are reasonably
expected to occur. Exposure distributions were then created using Monte Carlo
simulation with 100,000 iterations and the Latin Hypercube sampling method.
• The OES, use of paints and coatings - spray application, had the highest occupational
exposure for inhalation and dermal exposure; this is due to mist being generated during
application as well as a higher dermal loading value:
o Inhalation exposure for use of paints and coatings - spray application ranges from 5.5
mg/m3 (95th percentile, 8-hr TWA, resin-based paints) to 1.7/10 1 mg/m3 (50th percentile,
8-hr TWA, water-based paints). EPA identified mist generation as the main driver of
exposure but is not expected to occur during other COUs/OESs.
o Dermal acute retained dose (mg/kg-day) ranges from 8.02 (95th percentile) to 1.48 (50th
percentile).
The following subsections briefly describe EPA's approach to assessing occupational exposures and
results for each condition of use assessed. For additional details on development of approaches and
results refer to the Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental
Information File: Supplemental Information on Environmental Release and Occupational Exposure
Assessment (U.S. EPA 2024n). As discussed in Section 3.1.1, EPA has mapped the industrial and
commercial COUs to OESs in Table 3-1.
5.1.1.1 Approach and Methodology
As described in the Final Scope of the Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP)
CASRN115-96-8 (U.S. EPA 2020b). for each COU, EPA distinguishes exposures for workers and
ONUs. Normally, a primary difference between workers and ONUs is that workers may handle TCEP
and have direct contact with the chemical, while ONUs are working in the general vicinity of workers
but do not handle TCEP and do not have direct contact with it. Where possible, for each COU, EPA
identified job types and categories for workers and ONUs.
As discussed in Section 3.1.1, EPA established OESs to assess the exposure scenarios more specifically
within each COU. Table 3-1 provides a crosswalk between COUs and OESs. Figure 5-1 provides the
approaches used by EPA to estimate exposures for the OESs included in this risk evaluation of TCEP.
EPA did not identify any relevant inhalation exposure monitoring data to TCEP vapor for any of the
OESs, because TCEP does not have an Occupational Safety and Health Act (OSHA) permissible
exposure limit (PEL). For two OESs, monitoring data were available for TCEP in dust. The quality of
Page 163 of 638
-------
the monitoring data was evaluated using the data quality review evaluation metrics and the categorical
ranking criteria described in the 2021 Draft Systematic Review Protocol (U.S. EPA. 2021a). Relevant
data were assigned an overall quality determination of high, medium, low, or uninformative. In addition,
EPA established an overall confidence for the data when integrated into the occupational exposure
assessment. The Agency considered the assessment approach, the quality of the data and models, as well
as uncertainties in assessment results to assign an overall confidence level of robust, moderate, or slight.
Where monitoring data were reasonably available, EPA used these data to characterize central tendency
and high-end inhalation exposures. Where no inhalation monitoring data were identified, but inhalation
exposure models were reasonably available, EPA estimated central tendency and high-end exposures
using only modeling approaches. If both inhalation monitoring data and exposure models were
reasonably available, where applicable, EPA presented central tendency and high-end exposures using
both. EPA only identified measured dermal exposure estimates for dust generated at e-waste recycling
facilities. Monitoring data were not reasonably available for any other COUs. EPA standard models,
such as the EPA Mass Balance Inhalation Model and Fractional Absorption Model, were used to
estimate high-end and central tendency inhalation and dermal exposures for workers in each OES.
For many cases, EPA did not have monitoring data to estimate inhalation exposure for ONUs. In some
cases, this was addressed with the use of exposure models, when available. However, most OESs do not
contain inhalation exposure estimates for ONUs. In general, EPA expects ONU exposures to be less
than worker exposures. Dermal exposure for ONUs was not evaluated because these employees are not
expected to be in direct contact with TCEP.
Figure 5-1. Approaches Used for Each Component of the Occupational Assessment for Each OES
CDR = Chemical Data Reporting; GS = Generic Scenario; ESD = Emission Scenario Document; BLS = Bureau
of Labor Statistics; NIOSH (HHE) = National Institute for Occupational Safety and Health (Health Hazard
Evaluations); Fab = Fractional Absorption Model
In Table 5-1, EPA provides a summary for each OES by indicating whether monitoring data were
reasonably available; how many data points were identified, the quality of the data; EPA's overall
confidence in the data; whether the data were used to estimate inhalation exposures for workers and
ONUs; and whether EPA used modeling to estimate inhalation exposure to dust, vapors, or mist and
dermal exposures for workers and ONUs.
Page 164 of 638
-------
Table 5-2 provides a summary of EPA estimates for the total number of potentially exposed workers and
ONUs for each OES. To prepare these estimates, EPA first attempted to identify NAICS codes
associated with each OES. For these NAICS codes, EPA then reviewed Standard Occupational
Classification (SOC) codes from the Bureau of Labor Statistics (BLS) and classified relevant SOC codes
as workers or ONUs. All other SOC codes were assumed to represent occupations where exposure is
unlikely. EPA also estimated the total number of facilities associated with the NAICS codes previously
identified based on data from the U.S. Census Bureau.
EPA then estimated the average number of workers and ONUs potentially exposed per generic site by
dividing the total number of workers and ONUs by the total number of facilities. Finally, using EPA's
estimates for the number of facilities using TCEP, the Agency was able to estimate the total number of
workers and ONUs potentially exposed to TCEP for each OES. Additional details on EPA's approach
and methodology for estimating the number of facilities using TCEP and the number of workers and
ONUs potentially exposed to TCEP can be found in the Risk Evaluation for Tris(2-chloroethyl)
Phosphate (TCEP) - Supplemental Information File: Supplemental Information on Environmental
Release and Occupational Exposure Assessment (U.S. EPA. 2024nY
Page 165 of 638
-------
Table 5-1. Summary for Each PES
OES
Inhalation Exposure
Dermal Exposure
Monitoring
Modeling
Inhalation Exposure
Confidence"
Monitoring
Modeling
Dermal Exposure
Confidence"
Worker
# Data
Points
ONU
# Data
Points
Overall
Quality
Determ-
ination
Worker
ONU
Worker
ONU
Worker
Overall
Quality
Determ-
ination
Worker
Worker
ONU
Manufacture (import) -
Repackaging
X
N/A
X
N/A
N/A
V
X
Moderate
Slight
X
N/A
V
Moderate
N/A
Processing -
Incorporation into paints
and coatings - 1 -part
coatings
X
N/A
X
N/A
N/A
V
X
Moderate
Slight
X
N/A
V
Moderate
N/A
Processing -
Incorporation into paints
and coatings - 2-part
reactive coatings
X
N/A
X
N/A
N/A
V
X
Moderate
Slight
X
N/A
V
Moderate
N/A
Processing - Formulation
of TCEP-containing
reactive resins (for use in
2-part systems)
X
N/A
X
N/A
N/A
V
X
Moderate
Slight
X
N/A
V
Moderate
N/A
Processing - Processing
into 2-part resin article
X
N/A
X
N/A
N/A
V
X
Moderate
Slight
X
N/A
V
Moderate
N/A
Processing - Recycling e-
waste
V
55
~
21
High
X
X
Moderate
Moderate
X
N/A
Moderate
N/A
Distribution -
Distribution in commerce
Distribution in Commerce4
Industrial use -
Installation of article
1 (Surrogate)
X
N/A
High
X
X
Slight
Slight
X
N/A
X
N/A
N/A
Commercial use - Use
and/or maintenance of
articles
V
1 (Surrogate)
X
N/A
High
X
X
Slight
Slight
X
N/A
X
N/A
N/A
Commercial
use - Use of paints and
coatings - Spray
application
V
Surrogate
Spray GS
X
N/A
High
X
X
Moderate
Slight
X
N/A
V
Moderate
N/A
Commercial use - Lab
chemical - Use of
laboratory chemicals
X
N/A
X
N/A
N/A
V
X
Robust
Moderate
X
N/A
V
Moderate
N/A
Commercial uses:
X
N/A
X
N/A
N/A
X
X
N/A
N/A
X
N/A
X
N/A
N/A
Page 166 of 638
-------
OES
Inhalation Exposure
Dermal Exposure
Monitoring
Modeling
Inhalation Exposure
Confidence"
Monitoring
Modeling
Dermal Exposure
Confidence"
Worker
# Data
Points
ONU
# Data
Points
Overall
Quality
Determ-
ination
Worker
ONU
Worker
ONU
Worker
Overall
Quality
Determ-
ination
Worker
Worker
ONU
Furnishing, cleaning,
treatment/care products
fabric and textile
products
• Foam seating and
bedding products
Construction, paint,
electrical, and metal
products
• Building/construction
materials - insulation
Building/construction
materials - wood and
engineered wood
products - wood resin
composites
Disposal
Evaluated as part of each OES as opposed to a standalone OES
Where EPA was not able to estimate ONU inhalation exposure from monitoring data or models, this was assumed equivalent to the central tendency experienced by workers for
the corresponding OES; dermal exposure for ONUs was not evaluated because they are not expected to be in direct contact with TCEP.
" Robust confidence suggests thorough understanding of the scientific evidence and uncertainties. The supporting weight of scientific evidence outweighs the uncertainties to the
point where it is unlikely that the uncertainties could have a significant effect on the hazard estimate.
Moderate confidence suggests some understanding of the scientific evidence and uncertainties. The supporting scientific evidence weighed against the uncertainties is reasonably
adequate to characterize hazard estimates.
Slight confidence is assigned when the weight of scientific evidence may not be adequate to characterize the scenario, and when the assessor is making the best scientific
assessment possible in the absence of complete information. There are additional uncertainties that may need to be considered.
4 Distribution in Commerce of TCEP consists of the transportation associated with the moving of sealed containers of TCEP or sealed packages of TCEP containing products
(Section 3.1.1).
Page 167 of 638
-------
5.1.1.2 Summary of Inhalation Exposure Assessment
Table 5-2 summarizes the number of facilities and total number of exposed workers for all OESs.
Table 5-2. Summary of Total Number of Workers and ONUs Potentially Exposed to TCEP for
Each OES"
OES
Total Exposed
Workers/Site
Total
Exposed
ONUs/Site
Total Exposed/Site
(Exposure days/yr
High-End -
Central Tendency)
Number of
Facilities"
Notes
Manufacture
1
0
1
1 generic site
424690 - Other
(import) -
Repackaging
(7-4)
Chemical and Allied
Products Merchant
Wholesalers
Processing -
Incorporation into
14
5
19
(38-6)
1 generic site
325510 - Paint and
Coating Manufacturing
paints and coatings
- 1-part coatings
Processing -
Incorporation into
14
5
19
(2-1)
1 generic site
325510 - Paint and
Coating Manufacturing
paints and coatings
- 2-part reactive
coatings
Processing -
Formulation of
27
12
39
(6-1)
1 generic site
325211 - Plastics
Material and Resin
TCEP-containing
reactive resins (for
Manufacturing
use in 2-part
systems)
Processing -
Processing into 2-
part resin article
75
64
139
(250 - 72)
1 generic site
326400 - Aerospace
Product and Parts
Manufacturing
Processing -
2
2
4
Unknown
562920 - Materials
Recycling e-waste
(250-250)
Recovery Facilities
Distribution - Distribution in commerce
Distribution in commerce6
Industrial use -
Installation of
75
64
139
(250-250)
1 generic site
326400 - Aerospace
Product and Parts
article
Manufacturing
Commercial use -
Use and/or
75
64
139
(250-250)
1 generic site
326400 - Aerospace
Product and Parts
maintenance of
articles
Manufacturing
3
0
3
811121 - Automotive
Body, Paint, and
Commercial and
Industrial use - Use
of paints and
Sites vary based
on multiple
throughput
Interior Repair and
Maintenance
coatings - Spray
application
4
0
4
(Exposure days based
on 1-, 2-, or 250-day
scenarios)
scenarios; see
Table 3-2
238320 - Painting and
Wall Covering
Contractors
Page 168 of 638
-------
OES
Total Exposed
Workers/Site
Total
Exposed
ONUs/Site
Total Exposed/Site
(Exposure days/yr
High-End -
Central Tendency)
Number of
Facilities"
Notes
Commercial Use -
Lab chemical - Use
of laboratory
chemicals
3
3
6
(220-214)
13 sites (1st
percentile)
6 sites (5th
percentile)
541380 - Testing
laboratories
541713 - Research and
development in
nanotechnology
541714 - Research and
development in
biotechnology (except
nanobiotechnology)
541715 - Research and
development in the
physical, engineering,
and life sciences
(except nanotechnology
and biotechnology)
621511 - Medical
Laboratories
Commercial Uses -
• Furnishing, cleaning,
treatment/care products
o Fabric and textile products
o Foam seating and bedding
products
• Building/construction materials
o Insulation
o Wood resin composites
Manufacturing and processing for
these COUs has ceased
N/A
Disposal
Evaluated as part of each OES as opposed to a standalone OES
11 EPA's approach and methodology for estimating the number of facilities using TCEP and the number of workers
and ONUs potentially exposed to TCEP can be found in the Risk Evaluation for Tris(2-chloroethyl) Phosphate
(TCEP) - Supplemental Information File: Supplemental Information on Environmental Release and Occupational
Exvosure Assessment (U.S. EPA. 2024n).
h Distribution in Commerce of TCEP consists of the transportation associated with the moving of sealed containers of
TCEP or sealed packages of TCEP containing products (Section 3.1.1).
A summary of inhalation exposure results based on monitoring data and exposure modeling for each
OES is presented for workers in Table 5-3 and Table 5-4, respectively. ONUs are presented in Table
5-5. These tables provide a summary of time-weighted average (TWA) inhalation exposure estimates as
well as acute exposure concentrations (AC), average daily concentrations (ADC), lifetime average daily
concentrations (LADC), and subchronic average daily concentration (SCADC). The ADC is used to
characterize risks for chronic non-cancer health effects whereas the LADC is used for chronic cancer
health effects. The SCADC represents repeated exposure for approximately 30 days and is used for
intermediate exposure scenarios. Additional details regarding AC, ADC, LADC, and SCADC
calculations along with EPA's approach and methodology for modeling inhalation exposure can be
found in Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File:
Supplemental Information on Environmental Release and Occupational Exposure Assessment (U.S.
EPA. 2024nY
Page 169 of 638
-------
Table 5-3. Summary of Inhalation Exposure Results for Workers Based on Monitoring Data for Each PES
OES
Inhalation Monitoring (Worker, ppm)
TWA
AC
ADC
LADC
SADC
High-End
Central
Tendency
High-End
Central
Tendency
High-End
Central
Tendency
High-End
Central
Tendency
High-End
Central
Tendency
Processing - Recycling e-waste
9.68E-04
1.00E-07
6.6E-04
6.80E-08
4.51E-04
4.66E-08
2.31E-04
1.85E-08
4.83E-04
4.99E-08
Industrial use - Installation of
article
1.3E-05
1.3E-05
8.8E-06
8.8E-06
6.5E-06
6.5E-06
3.1E-06
2.4E-06
6.5E-06
6.5E-06
Commercial use - Use and/or
maintenance of articles
1.3E-05
1.3E-05
8.8E-06
8.8E-06
6.5E-06
6.5E-06
3.1E-06
2.4E-06
6.5E-06
6.5E-06
Table 5-4. Summary of Inhalation Exposure Results for Workers Based on Exposure Modeling for Each PES
OES
Inhalation Modeling (Worker, mg/m3)
TWA (8-hr)
AC
ADC
LADC
SADC
High-End
Central
Tendency
High-End
Central
Tendency
High-End
Central
Tendency
High-End
Central
Tendency
High-End
Central
Tendency
Manufacture (import) -
Repackaging
4.1E-02
1.1E-02
2.8E-02
7.5E-03
3.1E-03
8.9E-05
1.2E-04
3.4E-05
3.7E-03
1.1E-03
Processing - Incorporation into
paints and coatings - 1-part
coatings
1.0E-01
1.7E-02
7.1E-02
1.1E-02
8.0E-04
1.9E-04
3.2E-04
7.3E-05
9.2E-03
2.2E-03
Processing - Incorporation into
paints and coatings - 2-part
reactive
4.0E-01
9.6E-02
2.7E-01
6.5E-02
7.9E-04
1.9E-04
3.1E-04
7.1E-05
9.6E-03
2.3E-03
Processing - Formulation of
TCEP-containing reactive
resins (for use in 2-part
systems)
4.1E-01
7.4E-02
2.8E-01
5.1E-02
8.4E-04
1.8E-04
3.3E-04
6.9E-05
1.0E-02
2.2E-03
Processing - Processing into 2-
part resin article
1.8E-02
3.4E-03
1.2E-02
2.3E-03
2.3E-03
3.9E-04
9.2E-04
1.5E-04
8.1E-03
1.6E-03
Page 170 of 638
-------
OES
Inhalation Modeling (Worker, mg/m3)
TWA (8-hr)
AC
ADC
LADC
SADC
High-End
Central
Tendency
High-End
Central
Tendency
High-End
Central
Tendency
High-End
Central
Tendency
High-End
Central
Tendency
Distribution - Distribution in
commerce
Distribution in Commerce of TCEP consists of the transportation associated with the moving of sealed containers of TCEP or
sealed packages of TCEP containing products (Section 3.1.1).
Commercial and Industrial use
- Paints and coatings - Spray
(1-part coatings, 1-day
application)
(OES #7)
1.1
1.7E-01
7.5E-01
1. IE—01
2.1E-03
3.1E-04
1.1E-03
1.3E-04
2.5E-02
3.8E-03
Commercial and Industrial use
- Paints and coatings - Spray
(1-part coatings, 2-day
application)
1.1
1.7E-01
7.5E-01
1. IE—01
4.1E-03
6.3E-04
2.1E-03
1.37E-04
5.0E-02
7.7E-03
Commercial and Industrial use
- Paints and coatings - Spray
(1-part coatings, 250-day
application)
1.1
1.7E-01
7.5E-01
1. IE—01
5.1E-01
7.9E-02
2.6E-01
3.1E-02
5.5E-01
8.4E-02
Commercial and Industrial use
- Paints and coatings - Spray
(2-part coatings, 1-day
application)
5.5
8.5E-01
3.8
5.7E-01
1.0E-02
1.6E-03
5.3E-03
6.3E-04
1.3E-01
1.9E-02
Commercial and Industrial use
- Paints and coatings - Spray
(2-part coatings, 2-day
application)
5.5
8.5E-01
3.8
5.7E-01
2.1E-02
3.1E-03
1.1E-02
1.3E-03
2.5E-01
3.8E-02
Commercial and Industrial use
- Paints and coatings - Spray
(2-part coatings, 250-day
application)
5.5
8.5E-01
3.8
5.7E-01
2.6
3.9E-01
1.3
1.6E-01
2.8
4.2E-01
Commercial and Industrial use
- Lab chemical - Use of
laboratory chemicals
9.3E-04
5.8E-04
7.9E-04
5.1E-04
4.3E-04
2.7E-04
1.5E-04
8.8E-05
4.6E-04
2.9E-04
Disposal
Assessed as part of each OES and not as a stand-alone OES
Page 171 of 638
-------
Table 5-5. Summary of Inhalation Exposure Results for ONUs Based on Monitoring Data and Exposure Modeling for Each PES
OES
Inhalation Monitoring (ONU, mg/m3)
TWA
AC
ADC
LADC
SADC
High-End
Central
Tendency
High-End
Central
Tendency
High-End
Central
Tendency
High-End
Central
Tendency
High-End
Central
Tendency
Recycling of e-waste
1.9E-04
1.0E-07
1.3E-04
6.8E-08
8.9E-05
4.7E-08
4.5E-05
1.9E-08
9.5E-05
5.0E-08
Note: For many cases, EPA was not able to estimate inhalation exposure for ONUs, but EPA expects these to be lower than inhalation exposure for workers.
Page 172 of 638
-------
5.1.1.3 Summary of Dermal Exposure Assessment
Table 5-6 presents the estimated dermal acute retained dose for workers in various exposure scenarios.
The exposure estimates are provided for each OES based on the maximum possible exposure
concentration (Yderm), which is the highest concentration level of TCEP that a worker handles
throughout the process. The exposure concentration is determined based either on EPA's review of
currently available products and formulations containing TCEP or the assumption that neat TCEP is
handled to formulate these products.
The occupational dermal dose estimates assume one exposure event (applied dose) per workday and that
absorption through and into the skin may occur for up to 8 hours as representative of a typical workday.
Also, it is assumed that workers will thoroughly wash their hands with soap and water at the end of their
shifts. Regarding material remaining in the skin post-washing, EPA considers the quantity of material
remaining in the skin as potentially absorbable in accordance with OECD Guidance Document 156
(OECD, 2022). Therefore, overall occupational dermal exposure consists of the amount absorbed during
the 8-hour workday plus the amount remaining in the skin after washing the hands at the end of the 8-
hour workday.
In order to estimate occupational dermal exposures to TCEP, EPA relied on fractional absorption data
from Abdallah et al. (2016). This study used a low concentration (-0.005 wt percent in acetone) of
TCEP for in vitro dermal absorption testing of a finite dose (i.e., 500 ng/cm2) over a 24-hour period. As
mentioned above, the occupational exposure estimates are based on a typical 8-hour workday.
Cumulative absorption data from Abdallah et al. (2016) show 82.69 ng/cm2 absorbed after 8 hours of
exposure and the fraction remaining in the skin is 0.068 after 24 hours of exposure. Because there were
no data for the quantity remaining in the skin after 8 hours of exposure, EPA conservatively assumed
that the quantity in the skin after 24 hours of exposure is representative of the amount remaining in the
skin after 8 hours of exposure. EPA used the cumulative absorption data to determine the fraction
absorbed after an 8-hour exposure period (0.165), and then conservatively added the fraction remaining
in the skin at 24 hours (0.068). Therefore, the overall fractional absorption from an 8-hour exposure was
calculated for a dilute solution containing TCEP as Fabs = 0.165 + 0.068 = 0.233.
Page 173 of 638
-------
Table 5-6. Summary of Dermal Retained Dose for Workers Based on Exposure Modeling for Each
OES
OES
Max TCEP
Weight Fraction
(Max Yderm)
Non-occluded Worker Dermal Retained Dose
Dose (mg/day)
High-End
Central Tendency
Manufacture (import) - Repackaging
1.0
6.54
2.18
Processing - Incorporation into paints
and coatings - 1-part coatings
1.0
6.54
2.18
Processing - Incorporation into paints
and coatings - 2-part reactive coatings
1.0
6.54
2.18
Processing - Formulation of TCEP -
containing reactive resins (for use in 2-
part systems)
1.0
6.54
2.18
Processing - Processing into 2-part
resin article
4.0E-01
2.62
8.73E-01
Processing - Recycling e-waste
1.40E-05
4.4E-05
1.8E-05
Distribution - Distribution in commerce
Distribution in commerce 11
Industrial use - Installation of article
N/A
N/A
N/A
Commercial use - Use and/or
maintenance of articles
N/A
N/A
N/A
Commercial and Industrial use - Use of
paints and coatings - Spray application
OES
0.25
8.02
1.48
Commercial use - Lab chemical - Use
of laboratory chemicals
1.0
6.54
2.18
Commercial uses:
• Furnishing, cleaning,
treatment/care products
o Fabric and textile products
o Foam seating and bedding
products
• Construction, paint, electrical, and
metal products
o Building/construction
materials - insulation
o Building/construction
materials - Wood and
engineered wood products -
Wood resin composites
N/A
N/A
N/A
Disposal
Evaluated as part of each OES as opposed to a standalone OES
All dermal exposure scenarios are considered to be to a finite dose; therefore, no scenario is considered occluded.
11 Distribution in Commerce of TCEP consists of the transportation associated with the moving of sealed containers of
TCEP or sealed packages of TCEP containing products.
5.1.1.4 Weight of Scientific Evidence Conclusions for Occupational Exposure
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File:
Supplemental Information on Environmental Release and Occupational Exposure Assessment (U.S.
Page 174 of 638
-------
EPA. 2024n) provides a summary of EPA's overall confidence in its inhalation exposure estimates for
each of the OESs assessed.
5.1.1.4.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for
the Occupational Exposure Assessment
Number of Workers
Several uncertainties surround the estimated number of workers potentially exposed to TCEP. Current
CDR data reported in 2020 do not show production volumes that exceed the threshold of 25,000 pounds
and therefore, information was not available to estimate the number of workers associated with
manufacturing, processing, or use of TCEP.
There are inherent limitations to the use of CDR data as reported by manufacturers and importers of
TCEP. Manufacturers and importers are only required to report if they manufactured or imported more
than 25,000 lb of TCEP at a single site during any calendar year; as such, CDR may not capture all sites
and workers associated with any given chemical because it is possible for entities to use less than the
CDR threshold. Therefore, EPA assumes that any ongoing manufacturing, import, processing, or use of
TCEP occurs using volumes below the CDR threshold.
There are also uncertainties with BLS data, which are used to estimate the number of workers for the
remaining COUs. First, BLS' OES employment data for each industry/occupation combination are only
available at the 3-, 4-, or 5-digit NAICS level, rather than the full 6-digit NAICS level. This lack of
granularity could result in an overestimate of the number of exposed workers if some 6-digit NAICS are
included in the less granular BLS estimates but are not likely to use TCEP for the assessed applications.
EPA addressed this issue by refining the OES estimates using total employment data from the U.S.
Census' Statistics of U.S. Businesses (SUSB). However, this approach assumes that the distribution of
occupation types (SOC codes) in each 6-digit NAICS is equal to the distribution of occupation types at
the parent 5-digit NAICS level. If the distribution of workers in occupations with TCEP exposure differs
from the overall distribution of workers in each NAICS, then this approach will result in inaccuracy but
would be unlikely to systematically either overestimate or underestimate the count of exposed workers.
Second, EPA's judgments about which industries (represented by NAICS codes) and occupations
(represented by SOC codes) are associated with the uses assessed in this report are based on EPA's
understanding of how TCEP is used in each industry. Designations of which industries and occupations
have potential exposures is nevertheless subjective, and some industries/occupations with few exposures
might erroneously be included, or some industries/occupations with exposures might erroneously be
excluded. This would result in inaccuracy but would be unlikely to systematically either overestimate or
underestimate the count of exposed workers.
Analysis of Exposure Monitoring Data
This risk evaluation uses existing worker exposure monitoring data to assess exposure to TCEP during
some COUs, depending on availability of data. To analyze the exposure data, EPA categorized each data
point as either "worker" or "occupational non-user." The categorizations are based on descriptions of
worker job activity as provided in literature and EPA's judgment. In general, samples for employees that
are expected to have the highest exposure from direct handling of TCEP are categorized as "worker" and
samples for employees that are expected to have the lower exposure and do not directly handle TCEP
are categorized as "occupational non-user."
Exposures for ONUs can vary substantially. Most data sources do not sufficiently describe the proximity
of these employees to the TCEP exposure source. As such, exposure levels for the "occupational non-
Page 175 of 638
-------
user" category will have high variability depending on the specific work activity performed. It is
possible that some employees categorized as "occupational non-user" have exposures similar to those in
the "worker" category depending on their specific work activity pattern.
Some scenarios have limited exposure monitoring data in literature, if any. Where there are few data
points available, it is unlikely the results will be representative of worker exposure across the industry.
In cases where there was no exposure monitoring data, EPA used monitoring data from similar COUs as
a surrogate. For example, EPA did not identify inhalation monitoring data for installation of articles
based on the systematic review of literature sources. However, EPA estimated inhalation exposures for
this OES using monitoring data for TCEP exposures during furniture manufacturing (Makinen et al..
2009). EPA expects that inhalation exposures during furniture manufacturing occur from handling or
contacting TCEP-containing products, which is comparable to inhalation exposures expected during
installation of TCEP-containing products for aircraft or aerospace applications as well as automotive
parts and replacement parts. While these COUs have similar worker activities contributing to exposures,
it is unknown that the results will be fully representative of worker exposure across different COUs.
Where sufficient data were reasonably available, the 95th and 50th percentile exposure concentrations
were calculated using reasonably available data. The 95th percentile exposure concentration is intended
to represent a high-end exposure level, while the 50th percentile exposure concentration represents a
typical exposure level. The underlying distribution of the data, and the representativeness of the
reasonably available data, are not known. Where discrete data were not reasonably available, EPA used
reported statistics (i.e., 50th and 95th percentile). Because EPA could not verify these values, there is an
added level of uncertainty.
EPA calculated ADC and LADC values assuming workers and ONUs are regularly exposed during their
entire working lifetime, which likely results in an overestimate. Individuals may change jobs during
their career such that they are no longer exposed to TCEP, and actual ADC and LADC values would be
lower than the estimates presented.
The following describe additional uncertainties and simplifying assumptions associated with use of this
modeling approach for TCEP:
• No OSHA PEL (Very Little Monitoring Data): While EPA has confidence in the models used, it
is possible that they may not account for variability of exact monitoring processes and practices
at an individual site.
• No 2020 CDR Reporters and Only One 2016 CDR Reporter (with No Downstream Details
Provided): Assumptions of an ongoing production volume of 2,500 and 25,000 lb per site-year
could overestimate actual amount of TCEP handled at a given site, thus overestimating actual
exposures and releases. Release and exposure information using the 25,000 lb per site-year is
provided in the Engineering Supplemental file.
Modeled Dermal Exposures
The Fractional Absorption Model is used to estimate dermal exposure to TCEP in occupational settings.
The model assumes a fixed fractional absorption of the applied dose; however, fractional absorption
may be dependent on skin loading conditions. The model also assumes a single exposure event per day
based on existing framework of the EPA/OPPT 2-Hand Dermal Exposure to Liquids Model and does
not address variability in exposure duration and frequency. Additionally, the studies used to obtain the
underlying values of the quantity remaining on the skin (Qu) did not take into consideration the fact that
liquid retention on the skin may vary with individuals and techniques of application on and removal
from the hands. Also, the data used were developed from three kinds of oils; therefore, the data may not
Page 176 of 638
-------
be applicable to other liquids. Based on the uncertainties described above, EPA has a moderate level of
confidence in the assessed baseline exposure (see Table 5-1).
5.1.2 Consumer Exposures
TCEP - Consumer Exposures (Section 5.1.2):
Key Points
EPA evaluated the reasonably available information for the following consumer exposures, the key
points of which are summarized below:
• Limited information is available on TCEP in consumer products.
0 There are no current safety data sheets.
0 Weight fraction estimates in some cases were derived from literature values that were
over 20 years old and from maximum values reported in Washington State databases.
• The highest exposure estimates were from inhalation of the roofing insulation scenario (1.42
mg/kg/d) and the wood flooring scenario (1.24 mg/kg/day). However, EPA's confidence in
these estimates is low. Of the scenarios with moderate or robust confidence, the highest
inhalation and oral exposure estimates were from the textile for children's outdoor play
structures scenario (0.0604 mg/kg/day, 0.185 mg/kg/day, respectively).
• Inhalation is the driver for exposure to building and construction materials (e.g., roofing
insulation, acoustic ceiling) and wood flooring for adults.
• Oral ingestion is the driver for exposure for fabric and textile products, foam seating and
bedding products, and wooden television stands for children and infants.
5.1.2.1 Approach and Methodology
The migration of additive flame retardants from indoor sources such as building materials, fabrics,
textiles, and wood articles (from either ongoing COUs or in service products/articles at the end of their
life cycle) appear to be a likely source of flame retardants found in indoor dust, suspended particles, and
indoor air (Dodson et al.. 2012; Weschler and Nazaroff 2010). However, the relative contribution of
different sources of TCEP in these matrices is not well characterized. For example, building insulation,
textiles, and paints and coatings that contain TCEP have differing magnitudes of emissions that depend
on a variety of differing conditions.
Modeling was conducted to estimate exposure from the identified consumer COUs. Exposures via
inhalation, oral, and dermal routes to TCEP-containing consumer products were estimated using EPA's
Consumer Exposure Model (CEM), Version 3.2 (U.S. EPA 2023). Figure 5-2 below displays the
embedded models within CEM 3.2.
Page 177 of 638
-------
— Partition to Air
AJNH1:
Inhalation of
Airborne
Emissions from
Articles (Gas &
Particulate)
ARTICLE SOURCE
r
AJNGl:
Ingestion of
Airborne
Emissions from
Articles
(Particulates)
A_DER1: Dermal
Absorption of
Airborne
Emissions from
Articles
Aggregate
Total
Indoor Dose
by Pathway
and
Receptor
HUMAN RECEPTOR
Figure 5-2. Consumer Pathways and Routes Evaluated in this Assessment
CEM 3.2 estimates acute dose rates and chronic average daily doses for inhalation, ingestion, and
dermal exposures of consumer products and articles. CEM 3.2 gives exposure estimates for various
lifestages, including the following:
Adult
(>21 years)
Youth 2
(16-20 years)
Youth 1
(11-15 years)
Child 2
(6-10 years)
Child 1
(3-5 years)
Infant 2
(1-2 years)
Infant 1
(<1 year)
Lifetime LADD/LADC (lifetime average daily dose/lifetime average daily concentration)
Exposure inputs for these various lifestages are provided in EPA's CEM, Version 3.0 Appendices (U.S.
EPA. 2019d). CEM, 3.2 acute exposures are for an exposure duration of 1 day, and chronic exposures
are for an exposure duration of 1 year. For more information on specific use patterns, and exposure
inputs for populations, please see Appendix J A summary of key parameters used for the various
consumer exposures scenarios are provided in Table 5-10.
5.1.2.2 Consumer COUs and Exposure Scenarios
Page 178 of 638
-------
Table 5-7. Summary of Consumer CPUs, Exposure
Scenarios, and Exposure Routes
Life Cycle
Stage
Category
Subcategory
Consumer Use and Exposure
Scenario
Form(s)
Routes Evaluated
Consumer User
Oral
Inhalation
Dermal
Consumer
Use
Paints and
coatings
Paints and coatings
N/A
Liquid
Q
Vapor
Q
Mist
Q
Consumer
Use
Furnishing,
cleaning,
treatment/care
products
Fabric and textile products
Direct contact through use of
products/articles containing
TCEP
Air/Particulate
~
Dust
~
~
Article/Product
C ontact/Mouthing
~
~
Consumer
Use
Furnishing,
cleaning,
treatment/care
products
Foam seating and bedding
products
Direct contact through use of
products/articles containing
TCEP
Air/Particulate
~
Dust
~
~
Article/Product
C ontact/Mouthing
~
~
Consumer
Use
Construction,
paint,
electrical, and
metal products
Building/construction
materials - insulation
Direct contact through use of
building/construction materials
made containing TCEP
Air/Particulate
~
Dust
~
~
Article/Product
Contact'1
Building/construction
materials - Wood and
engineered wood products -
Wood resin composites
Direct contact through use of
wood and wood products made
containing TCEP
Air/Particulate
~
Dust
~
~
Article/Product
C ontact/Mouthing
~
~
Disposal
Wastewater,
liquid wastes,
and solid
wastes
Wastewater, liquid wastes,
and solid wastes
Direct contact through use of
products/articles containing
TCEP
Article/Product Contact
Q
Dust
Q
Air/Particulate
Q
Long-term emission/mass-
transfer through use of products
containing TCEP
Dust
Q
Air/Particulate
Q
Quantitatively assessed; Q = Qualitatively assessed
11 Contact with the product is not expected (see Section 5.1.2.2.1).
Page 179 of 638
-------
Paints and Coatings - Including Those Found on Automotive Articles and Replacement Parts
Consumers are no longer able to purchase paints and coatings containing TCEP because their domestic
retail production and manufacturing has ceased. It is possible that old paint cannisters stored in
basements, crawlspaces, and/or garages may result in exposure to TCEP from off-gassing or during use
by consumers.
Furthermore, the exposure to paints and coatings containing TCEP may occur via an article scenario in
which the paint and coating has already been applied. There is a higher likelihood that older buildings
and vehicles may have used TCEP-containing paints and coatings when the use of TCEP in consumer
paints and coatings was more common. This dried scenario is like the acoustic ceilings/drywall scenario
that was assessed for the building/construction materials COU. The exposure scenario of dried paints
and coatings present in the indoor environment is qualitatively assessed and described in Section
5.1.2.2.5.
Due to limited reasonably available information regarding the use of paints and coatings and the
uncertainties surrounding the weight fraction, activity and use patterns, and duration of use, EPA did not
quantitatively assess the use of paints and coatings- including those found on automotive articles and
replacement parts containing TCEP. See Section 5.3.2.2.2 for a qualitative assessment of consumer use
of paints and coatings.
Fabric and Textile Products
In a study of the CHAMACOS cohort in California, Castorina et al. (2017) indicates that TCEP levels in
dust are significantly associated with the presence of extremely worn carpets. Crowding, poor housing
quality, and lack of maintenance by landlords can result in "extremely worn" carpets, warranting
replacement. This suggests that individuals who are lower socioeconomic status may have increased
exposure to TCEP due to the inability to replace extremely worn carpets.
Ionas et al. (2014) measured TCEP concentrations in different types (e.g., hard plastic, soft plastic and
rubber, wood and foam and textile) of children's toys in Antwerp, Belgium. This study reported a
median TCEP concentration of 3 |ig/g, mean of 10 |ig/g, and maximum of 45 |ig/g of TCEP in 36
percent in 25 foam and textile products sampled. For soft plastics and rubber products, a detection
frequency of 42 percent in 31 toys with a median of 5 |ig/g, mean of 10 |ig/g, and maximum of 65 |ig/g
was reported. For hard plastic toys, the study author reported a detection frequency of 14 percent in 50
toys with a median of 2 |ig/g, mean of 10 |ig/g, and maximum of 25 |ig/g. These mean concentrations
correspond to a weight fraction of 0.001 percent.
EPA searched the Ecology Washington database (WSDE. 2023) in August 2022 and retrieved various
information for fabric and textile products containing TCEP. The Ecology Washington database
sampled for fabric and textile products that are likely to be mouthed or used by children under the age of
three. The database had 67 products classified as textiles (synthetic fibers and blends), there were 2
detects at 0.01 percent and 1.3 percent. The 1.3 percent weight fraction was detected in the surface
textile of a children's mini chair. The database indicated four detects of TCEP in carpet padding and rug
mats. The weight fractions for these carpet products ranged from 0.01 to 0.02 percent.
Little additional information was found in the literature search on the percentages of TCEP in carpet
back coating. A European patent has suggested that flame retardants may be generally used in carpet
back coating at between 5 to 30 percent (Herrlich et al.. 2013).
Page 180 of 638
-------
Two scenarios were modeled for the fabric, textile, and leather products not covered elsewhere—one for
an outdoor children's play structure and one for carpet back coating. The CEM 3.2 scenario used for
both scenarios were Fabrics: curtains, rugs, wall coverings (see Table 5-9). Values of 1.3 percent for
fabric in children's play structure and 0.02 percent for the carpet back coating were selected for weight
fractions for consumer modeling as these values are believed to be more representative of products
readily available in the United States.
Foam Seating and Bedding Products
Various studies have reported the use of TCEP in furniture, automotive, and bedding foams (Maddela et
al.. 2020). In the early 2000s, Ingerowski et al. (2001) recorded TCEP in mattresses at 890 mg/kg
(0.09%) in Germany. Ali et al. (2012) reported much lower concentrations of TCEP on mattresses
surfaces (0.11 ng/g) in New Zealand. Two different case reports reported the acute death of dogs (one
report on two pit bulls and one report on a German shepherd and a rottweiler) after chewing old
automobile foams. The case studies found significant amounts (>2 ppm) of TCEP in their stomach
contents (Lehner et al.. 2010).
Fang et al. (2013) has measured another flame retardant (V6) at levels of 3.63 percent in couch foam and
5.3 percent in auto foams. TCEP has been reported to be an impurity in V6 of up to 14 percent. V6 is the
dimer of TCEP, and it would be expected that TCEP would be an impurity of a V6 mixture. Hence, the
product of these two values suggests TCEP is available in couch foams at 0.51 percent and in auto
foams at 0.74 percent (Fang et al.. 2013).
Hoehn et al. (2024) sampled foam seats from 51 vehicles (model year of 2015 or newer) across the
United States. Only one sample detected of TCEP in auto foam however the authors did not report the
weight fraction or amount detected. TCEP was detected in silicone samplers up to 391 mg/kg in the
winter in 14 percent of vehicles, and up to 4,981 mg/kg in the summer in 44 percent of vehicles
suggesting that seasonal variations and increases in temperature may lead to more expulsion of TCEP
from materials inside a vehicle cabin (Hoehn et al.. 2024).
Ingerowski et al. (2001) recorded TCEP in polyurethane soft foam at 19,800 mg/kg (1.98%), values
from Fang et al. (2013) were selected for this furniture foam and auto foam scenarios as they were
thought to be more current and representative of the U.S. population.
Bradman et al. (2014) sampled indoor dust concentrations between childcare facilities with and without
foam napping equipment in 2011 in California. Median TCEP concentrations were significantly higher
in rooms with foam napping equipment (median of 642.9 ng/g) vs. rooms without foam napping
equipment (median of 260.9 ng/g).
For the foam toy block scenario, a weight fraction of 0.64 percent was calculated using information from
Fang et al. (2013). This was based on the knowledge of 4.6 percent of V6 in polyurethane foam with an
understanding that TCEP has been reported to be an impurity in V6 of up to 14 percent. Ionas et al.
(2014) reports a lower weight fraction (0.001%) of TCEP in 25 foam and textile toys.
Building/Construction Materials — Insulation
TCEP has been reportedly used in building materials, including wood preservations coatings, glass fiber
wallpapers, and acoustic ceilings (Maddela et al.. 2020). High TCEP concentrations in dust (94 mg/kg)
at a Swedish library were suggested to have been due the use of TCEP in the acoustic ceiling (Marklund
et al.. 2003).
Page 181 of 638
-------
Ingerowski et al. (20011 reported TCEP in polyurethane soft foam at 19,800 mg/kg (1.98%), and 68,000
mg/kg (6.8%) in acoustic ceilings. Kaiiwara et al. (2011) recorded concentrations of TCEP in insulation
boards of up to 10 ng/g in products purchased in Japan.
To assess the building/construction materials scenario, two exposure scenarios were run in GEM 3.2:
roofing insulation (under the Plastic articles - foam insulation scenario) and acoustic ceiling (under the
Drywall scenario). The weight fractions used for this modeling were 1.98 and 6.8 percent, respectively.
These exposures scenarios measured the chronic release of TCEP from the roofing insulation and
acoustic ceiling to the indoor air and indoor dust. They did not consider do-it-yourself (DIY) scenarios
of a consumer installing these articles because they are no longer commercially available.
Wood and Engineered Wood Products
A case study reported neurotoxic signs (muscular weakness) experienced by a 5-year-old child after
exposure to TCEP. It was postulated that the exposure was due to wood paneling that had been treated
with a wood preserver coating containing 3 percent TCEP. However, TCEP in dust was not quantified.
The study reported 600 mg/kg (0.06%) of TCEP in wood as cited in SCHER (2012). Jonas et al. (2014)
reported a detection frequency of 25 percent in 8 wooden toys with a median of 4 jag/g, mean of 4 ug/g,
and maximum of 5 |ig/g, which corresponds to a mean weight fraction of 0.0004 percent. The products
sampled in lonas et al. (2014) were around 2007, with around half of the products coming from China.
Anecdotally, TCEP concentrations have been reported to be present in imported wooden TV stands. The
photo below lists TCEP on a California Proposition 65 label on a wooden TV stand product imported to
the United States from Malaysia (Figure 5-3).
To assess the wood and engineered wood products scenario, two exposure scenarios for wood products
(exposure from wood flooring and wooden TV stand) was run in CEM 3.2 utilizing the wood articles:
hardwood floors, furniture predefined scenario with a weight fraction of 3 percent.
Wastewater, Liquid Wastes, and Solid Wastes
Consumers may be exposed to articles containing TCEP during the handling of disposal and waste. The
removal of articles in DIY renovation scenarios may lead to direct contact with articles and the dust
generated from the articles leading to consumer exposure. Due to the difficulties in quantifying
consumer disposal of products containing TCEP, consumer disposal of TCEP was not quantitatively
Figure 5-3. Photo of TCEP Label on Wooden Television Stand
Source: Photo by Yousuf Ahmad, U.S. EPA.
Page 182 of 638
-------
assessed for this risk evaluation. Section 5.1.2.2.5 discusses the qualitative assessment for consumer
disposals including the landfilling of building products and articles that contain TCEP.
5.1.2.2.1 Consumer Exposure Routes Evaluated
The COUs that were evaluated for TCEP were all articles. As such, the relevant underlying models
utilized for TCEP included those listed in Table 5-8 below.
Table 5-8. CEM
3.2 Model Codes and Descriptions
Model Code
Description
E6
Emission from article placed in environment
AINHl
Inhalation from article placed in environment
AING1
Ingestion after inhalation
AING2
Ingestion of article mouthed
AING3
Incidental ingestion of dust
ADERl
Direct transfer from vapor phase to skin
ADER2
Dermal dose from article where skin contact occurs
ADER3
Dermal dose from skin contact with dust
CEM 3.2 contains 73 specific product and article categories and several generic categories that can be
user-defined for any product and article. Table 5-9 presents a crosswalk between the COU subcategories
with these predefined scenarios. In some cases, one COU mapped to multiple scenarios, and in other
cases one scenario mapped to multiple COUs.
Table 5-9. Crosswalk of COU Subcategories, CEM 3.2 Scenarios, and Relevant CEM 3.2 Models
Used for Consumer Modeling
TCEP COU Subcategory
Exposure Scenario
CEM 3.2 Scenario
c\
Z
X
z
o
z
o
z
o
e
m
&
Carpet back coating
Fabrics: curtains, rugs,
wall coverings
Fabric and textile products
Textile for outdoor
children's outdoor
play structures
Fabrics: curtains, rugs,
wall coverings
Foam seating and bedding
product
Foam used in
automobiles, foam
used in living room
furniture
Plastic articles: furniture
(sofa, chairs, tables)
Mattress
Plastic articles: mattresses
Other foam objects
(toy blocks)
Plastic articles: other
objects with potential for
routine contact (toys,
foam blocks, tents)
Page 183 of 638
-------
>
>
>
>
TCEP COU Subcategory
Exposure Scenario
CEM 3.2 Scenario
C\
HH
Z
X
1—'
HH
Z
o
1—'
HH
Z
o
HH
Z
o
*3
i—'
*3
*3
Building/construction
materials - insulation
Insulation
Plastic articles: foam
insulation
•
•
•
•
•
•
Acoustic ceiling
Drywall (acoustic ceiling)
•
•
•
•
•
•
Building/construction
materials - wood and
Wood flooring
Wood articles: hardwood
floors, furniture
engineered wood products
- wood resin composites
Wooden TV stand
Wood articles: hardwood
floors, furniture
In total, the four COUs for TCEP were mapped to nine CEM 3.2 scenarios. Relevant consumer
behavioral pattern data (i.e., use patterns) and product-specific characteristics were applied to each of
the scenarios. For more information on specific use patterns and product-specific characteristics please
see Appendix J.
Inhalation, oral and dermal routes were evaluated for each of the article COUs. The article model
Ingestion of article mouthed (A ING2) was only evaluated for the COUs where it was anticipated that
mouthing of the product would occur. For example, it is unlikely that a child will mouth roofing
insulation or an acoustic ceiling, hence the A ING2 Model was deemed inappropriate for estimating
exposure for these COUs. The ADER2 Model (dermal dose from article where skin contact occurs)
was not used for estimating dermal exposure to roofing insulation and acoustic ceilings because dermal
contact is not expected to occur for these articles.
The chronic and lifetime exposure estimates are the most relevant durations for consumer articles.
Furnishings, building materials, and foam seating and bedding products are typically used over a longer
time frame than other types of consumer products with direct applications (e.g., household cleaners,
solvents). The exposure scenario of relevance for consumers for building and construction materials,
fabric and textile products, and foam seating and bedding products is that of a repeated exposure over a
chronic duration. As such, the exposure estimates presented in the successive sections focus on the
chronic average daily doses rather than the acute estimates. A summary of the acute, chronic, and
lifetime exposure estimates are presented in Section 5.1.2.3 and further discussed in Appendix 1.5.6.
The CEM, Version 3.2 was selected for the consumer exposure modeling as the most appropriate model
to use based on the type of input data available for TCEP-containing consumer products. The advantages
of using CEM to assess exposures to consumers and bystanders are as follows:
• CEM model has been peer-reviewed;
• CEM accommodates the distinct inputs available for the products containing TCEP; and
• CEM uses the same calculation engine to compute indoor air concentrations from a source as the
higher-tier Multi-Chamber Concentration and Exposure Model (MCCEM) but does not require
measured chamber emission values (which are not available for TCEP).
Consumer modeled exposure estimates were compared to the reported monitoring and reported modeled
estimates for indoor air and indoor dust. Residential indoor air, indoor dust, and personal breathing zone
data were identified and evaluated during systematic review (U.S. EPA 2024r. x). Sections 3.4.1 and
3.4.2 provide a summary of the reported monitoring and reported modeled data in indoor air and indoor
dust. A challenge in comparing EPA modeled exposures estimates with the reported monitoring and
Page 184 of 638
-------
modeled data in the literature is that EPA's modeled exposure estimates are by COU, whereas reported
information in the literature are not typically specified by COU. For a characterization of model
sensitivity and full exposure results, see Appendix J.
Page 185 of 638
-------
Table 5-10. Summary of Key Parameters for Article Modeling in CEM 3.2"
Consumer
Exposure
Scenarios
Initial
Concentration of
SVOC in Article
(mg/cm3)
Weight
Fraction of
Chemical
(%)
Density
Product/Article
(g/cm3)
Duration of
Article
Contact (min)
Frequency of
Article Contact
(Events/Day)
Surface
Area of
Article (m2)
Thickness
of Article
Surface
Layer (m)
Interzone
Ventilation
Rate (m3/h)
Use
Environment
Volume (m3)
Textile-
outdoor
play
structures
4.03
1.30
0.31
180
1
17.8608
0.055
1E-30
492
Carpet back
coating
4.00E-02
0.02
0.2
1,140
5
1.6
0.5
1E-30
492
Foam living
room
2.22E01
0.74
0.03
600
10
0.4225
0.01
88.6092
50
Foam auto
2.22E01
0.74
0.03
600
1
0.4225
0.01
9.4872
2.4
Mattresses
2.67E-02
0.09
0.03
600
1
3.097
0.5
107.01
36
Other foam
objects
1.92E-01
0.64
0.03
3.8
40
0.6606
0.01
108.978
50
Roofing
insulation
5.94E-01
1.98
0.03
0
1
158
0.5
1E-30
492
Wood
flooring
3.00E01
3.00
1
1,140
10
211
0.1
88.6092
50
Wood TV
stand
3.00E01
3.00
1
120
10
1.38
0.1
88.6092
50
Acoustic
ceiling
1.12E01
6.80
0.165
0
1
12.6
0.5
107.01
36
"For detailed information on selection of parameters refer to Risk Evaluation for Tris(2-chloroethvl) Phosphate (TCEP) - Supplemental Information File: Consumer
Exposure Modeling Inputs (U.S. EPA, 2024e).
Page 186 of 638
-------
5.1.2.2.2 Inhalation Exposure Assessment
Due to its vapor pressure of 0.0613 mm Hg at 25 °C, it is expected that under non-heated conditions
TCEP concentrations in air would be negligible. However, research has indicated that inhalation
exposure of TCEP can be higher than dermal exposure (Ortiz Carrizales. 2018). In addition,
concentrations of TCEP in the indoor air have been shown to be higher than ambient air concentrations
(Wong et al.. 2018). In general, concentrations of organophosphate flame retardants increase both
indoors and outdoors during warmer seasons (Wang et al.. 2019a).
Generally, TCEP release is higher at higher temperatures. However, the material to air coefficient (Kma)
values for TCEP have been shown to be similar at 35 and 55 °C. This implies that after reaching a
certain temperature, TCEP emission rates increase in a KMA-independent manner with further increase in
temperature. The Kma value at 23 °C for polyisocyanurate (PIR) foam was 7.76x 106 and for
polyurethane foam (PUF) was 3.87><106 (Maddela et al.. 2020).
Due to its presence in particulates both less than and greater than 2.5 pm, and its presence in the gaseous
phase, EPA expects both inhalation pathways (<2.5 pm deposits in lung and <0.1 pm deposits in
alveolar region) and ingestion pathways (>2.5 pm deposits in mouth) to be contributors to TCEP
exposure. See Section 3.3.1.2.1 for more details regarding the particle vs. gas phase distribution of
TCEP. Consumer inhalation exposure to TCEP is expected through the direct inhalation of indoor air
and dust. Table 5-11 below illustrates the steady state SVOC concentrations and respirable particle (RP)
concentrations resulting from consumer exposure to articles containing TCEP.
Table 5-11. Steady State Air Concentrations and Respirable Particle of TCEP from Consumer
Modeling (CEM 3.2)
COU Subcategory
Consumer Scenario
Air SVOC
(mg/m3)
Respirable
Particles
(Hg/mg)
Fabric and textile products
Carpet back coating
3.06E-02
3.79E-02
Textile-outdoor play
structures
3.97
4.81
Foam seating and bedding product
Foam auto
1.04E-04
2.43E-05
Foam living room
9.34E-06
3.33E-06
Mattresses
4.45E-04
1.33E-04
Other foam objects
1.26E-05
4.50E-06
Building/construction materials - insulation
Roofing insulation
9.32
1.13E01
Acoustic ceiling
7.52E-01
2.25E-01
Building/construction materials - wood and
engineered wood products - wood resin composites
Wood flooring
6.59
2.34
Wood TV stand
4.31E-02
1.53E-02
The insulation scenario followed by the wood-resin scenario had the highest TCEP air concentrations
(9.32 and 6.59 mg/m3, respectively).
Exposures doses (chronic average daily inhalation doses [CADDs]) for all of the COU subcategories
were estimated for the inhalation pathway via the following formulae) (A INHl):
Page 187 of 638
-------
Equation 5-1.
CADDAir —
C,
gas_avg
x FracTime x InhalAfter x CF1
Equation 5-2.
CADDParticuiate —
Equation 5-3.
SVOCRPn
X RP„
BW x CF-,
x (1 — IFRP)FracTime x InhalAfter x CFX
Where:
CADDAir
CADDParticuiate
CADDtotai
r
ugas_avg
SVOCRPairavg
RPair_avg
IFrp
FracTime
InhalAfter
CFi
BW
cf2
BW x CF?
CADDtotal — CADDAir + CAD Dparticuiate
Potential Chronic Average Daily Dose, air (mg/kg-day)
Potential Chronic Average Daily Dose, particulate (mg/kg-day)
Potential Chronic Average Daily Dose, total (mg/kg-day)
Average gas phase concentration (|ig/m3)
Average SVOC in RP concentration, air (|ig/mg)
Average RP concentration, air (mg/m3)
RP ingestion fraction (unitless)
Fraction of time in environment (unitless)
Inhalation rate after use (m3/hr)
Conversion factor (24 hr/day)
Body weight (kg)
Conversion factor (1,000 |ig/mg)
Exposures doses (Acute Dose rate ADRs) for all of the COU subcategories were estimated for the
inhalation pathway via the following formulae (A INHl):
Equation 5-4.
ADRAir =
Equation 5-5.
AD Rp articulate
Equation 5-6.
Cgas max x FracTime x InhalAfter x CFX
BW x CF2
SVOCRPair max x RPair_avg x FracTime x InhalAfter x CFt
BW x CF2
Where:
ADRAir
AD Rp articulate
ADRtotal
ADRfotal — ADRAir + ADR particulate
Potential Acute Dose Rate, air (mg/kg-day)
Potential Acute Dose Rate, particulate (mg/kg-day)
Potential Acute Dose Rate, total (mg/kg-day)
Page 188 of 638
-------
Cgas max = Maximum gas phase concentration (|ig/m3)
SVOCRPair max = Maximum SVOC in RP concentration, air (|ig/mg)
RPair_max = Maximum RP concentration, air (mg/m3)
FracTime = Fraction of time in environment (unitless)
InhalAfter = Inhalation rate after use (m3/hr)
CF1 = Conversion factor (24 hr/day)
BW = Body weight (kg)
CF2 = Conversion factor (1,000 |ig/mg)
The ADR and CADD equations (Equation 5-1, Equation 5-2, Equation 5-3, Equation 5-4, Equation 5-5,
and Equation 5-6) for A INHl consider both contributions from air and particulates. The average gas
phase concentration is considered for CADDair, and the maximum gas phase concentration is
considered for ADRair. The average SVOC in the RP concentration is considered for CADDparticulate,
and the maximum SVOC in the RP concentration is considered for ADRparticulate. CADDair and
CADDparticulate are summed to obtain CADDtotal. Similarly, ADRair and ADRparticulate are
summed to get ADRtotal. The SVOC in the RP concentration is given in |ig/mg and is multiplied by an
average RP concentration (in mg/m3).
Although the inhalation exposures to consumer articles containing TCEP are dominated by gas phase air
concentrations vs. the SVOC in RP concentrations, EPA decided to include both in the inhalation
exposure estimates. Therefore, EPA presented consumer inhalation values as doses (mg/kg-day), rather
than concentrations (mg/m3), because the dose values expressed as mg/kg-day included contributions
from both the gas and particulate phases.
CEM 3.2 outputs include inhalation doses for all lifestages. Inhalation doses are calculated for lifestages
by varying the BW and inhalation rate for the various population groups. These inhalation dose
calculations are simplified and do not take into consideration lifestages differences in ventilation,
anatomy, and metabolism. This risk evaluation presents one inhalation value (the adult value) by COU
(see Table 5-15 and Table 5-16). Appendix J. 1.3 presents the reported CEM inhalation doses with
breathing weight and body weight adjustments for all lifestages.
A summary of the acute, chronic, and lifetime inhalation doses are presented in Section 5.1.2.3. Table
5-10 presents a summary of the key parameters used for consumer modeling with CEM 3.2. For more
information on CEM 3.2, input parameters, sensitivity analysis, and assumptions used for consumer
modeling please see Appendix J.
5.1.2.2.3 Dermal Exposure Assessment
Consumers may be dermally exposed to TCEP via skin contact with consumer articles, skin contact with
dust generated from consumer articles, or the deposition of vapor generated from articles onto the skin.
CEM 3.2 contains dermal modeling components that estimate absorbed dermal doses resulting from
dermal contact with chemicals found in consumer products: Direct transfer from vapor phase to skin
(A DERl), dermal dose from article where skin contact occurs (A DER2), and dermal dose from skin
contact with dust (A DER3). All three models were used to estimate exposure to articles containing
TCEP, except for A DER2, which was not used for the Building/construction materials - insulation
COU because direct article contact with skin was not expected.
Contact of skin with articles drives the dermal exposure estimate in cases where contact is expected.
Otherwise, skin contact with dust is the driver of dermal exposure. The following equation was used to
calculate CADD for A DER2:
Page 189 of 638
-------
Equation 5-7.
Where:
CADD
r
^art
SA/BW
FRabsart
EDcr
ATcr
1
CADD =
Si4
^art * X I X FRai)s art * EDcr
ATrr-
Potential Chronic Average Daily Dose (mg/kg-day)
Chemical concentration in article (mg/cm3)
Surface area to body weight ratio (cm2/kg)
Fraction absorbed (unitless)
Exposure duration, chronic (years)
Averaging time, chronic (years)
Average distance a diffusing molecule travels per contact (cm/day)
Many of these parameters are calculated within CEM 3.2. The parameter / is a function of duration of
article contact (min/day). ADER3 has a similar formula:
Equation 5-8.
CADD =
SA
Dustcr wgt x x AF x FA x EvD x EDcr
CF1 x ATC
Where:
Dust,
AF
FA
EvD
CFi
crjwgt
Chronic weighted dust concentration ([j,g/mg)
Adherence factor of dust to hand (mg/cm2-event)
Fraction absorbed (unitless)
Frequency of article contact per day (events/day)
Conversion factor (insert value)
Compared to ADER2, this formula substitutes a chronic weighted dust concentration for the chemical
concentration and replaces the / term with an adherence factor (AF) and frequency of article contact
(EvD).
A key parameter in estimating results for ADER2 and ADER3 is fraction absorbed (Fabs). While the
duration of interaction with materials that contain TCEP may be shorter than the duration that was tested
in the dermal absorption study (i.e., a 24-hour exposure), EPA cannot assume that consumers would
immediately wash their hands following contact with treated objects (e.g., carpets). Therefore, the dose
that is deposited on the skin during an activity would be expected to remain on the skin until the skin is
eventually washed. As a result, EPA applied a 24-hour value for fraction absorbed (35.1%) from
Abdallah et al. (2016) to all consumer dermal exposures scenarios.
Table 5-12 provides the chronic dermal doses from each of the underlying models in CEM 3.2 and for
adults and children 3 to 6 years of age. All lifestages were analyzed. For more information on the
consumer dermal exposure inputs, equations, results (for all lifestages) and sensitivity analysis please
see Appendix J and EPA's CEM 3.0 Appendices (U.S. EPA 2019d).
Page 190 of 638
-------
Table 5-12. Chronic Dermal Average Daily Doses (mg/kg-day) of TCEP from Consumer Article
Modeling for Adults and Children 3 to 6 Years of Age (CEM 3.2)
COU Subcategory
Consumer
Scenario
Lifestage
A DERI
Vapor to
Skin
ADER2
Skin
Contact
ADER3
Skin
Contact
with Dust
Total
Chronic
Dermal ADD
Fabric and textile
products
Carpet back
coating
Adult
2.29E-07
3.16E-04
3.45E-05
3.50E-04
Child
3.68E-07
5.07E-04
5.53E-05
5.63E-04
Textile -
outdoor play
structures
Adult
2.97E-06
1.26E-02
8.41E-04
1.35E-02
Child
4.78E-06
2.03E-02
1.35E-03
2.17E-02
Foam seating and
bedding product
Foam auto
Adult
3.88E-10
5.65E-03
1.78E-08
5.65E-03
Child
6.44E-10
9.38E-03
2.96E-08
9.38E-03
Foam living
room
Adult
6.95E-10
1.26E-02
2.16E-08
1.26E-02
Child
1.16E-09
2.10E-02
3.60E-08
2.10E-02
Mattresses
Adult
3.32E-08
1.54E-03
3.99E-07
1.54E-03
Child
5.51E-08
2.55E-03
6.64E-07
2.55E-03
Other foam
objects
Adult
9.40E-10
8.69E-04
1.16E-07
8.69E-04
Child
1.56E-09
1.44E-03
1.92E-07
1.44E-03
Building/construction
materials - insulation
Roofing
insulation
Adult
3.49E-05
0.00
9.98E-04
1.03E-03
Child
5.61E-05
0.00
1.61E-03
1.66E-03
Acoustic ceiling
Adult
5.64E-06
0.00
6.79E-05
7.36E-05
Child
9.05E-06
0.00
1.09E-04
1.18E-04
Building/construction
materials - wood and
engineered wood
products - wood
resin composites
Wood flooring
Adult
4.94E-05
2.37E-01
4.05E-03
2.41E-01
Child
7.93E-05
3.80E-01
6.51E-03
3.87E-01
Wood TV stand
Adult
3.23E-07
7.68E-02
2.65E-05
7.69E-02
Child
5.19E-07
1.23E-01
4.26E-05
1.23E-01
5.1.2.2.4 Oral Exposure Assessment
Consumers may be exposed to TCEP via transfer from hand to mouth, ingestion after inhalation, mouthing of
articles, and the incidental ingestion of dust generated from consumer articles. CEM 3.2 contains an
ingestion modeling component that estimates ingestion doses resulting from consumer products:
ingestion after inhalation (AING1), ingestion of article mouthed (AING2), and incidental ingestion
from dust (A ING3). All three models were used to estimate exposure to articles containing TCEP,
except for A ING2, which was not used for the building/construction materials COU as mouthing of the
article was not expected.
Mouthing is a particular important route for estimating exposure to children and infants who may have
higher exposures to toys and children's products. CEM 3.2 has four choices for mouthing scenarios: 0, 1
(low), 10 (medium), and 50 (high) cm2. The high mouthing input was selected for outdoor play
structures and other foams (toy blocks) because these are children's products. The medium values were
selected for carpet back coating, wood flooring, wooden TV stand, foam furniture in the living room,
foam seat in an automobile, and the mattress scenarios.
The following equation was used to calculate CADD for A ING2:
Page 191 of 638
-------
Equation 5-9.
Where:
CADD
MR
CA
Dm
EDcr
CFi
cf2
ATcr
BW
CADD =
MR x CAx Dmx EDcr x CF1
BW x ATcr x CF2
Potential Chronic Average Daily Dose (mg/kg-day)
Migration rate of chemical from article to saliva (mg/cm2/hr)
SA/BW= Surface area to body weight ratio (cm2/kg)
Duration of mouthing (min/hr)
Exposure duration, chronic (years)
Conversion factor (24 hr/day)
Conversion factor (60 min/hr)
Averaging time, chronic (years)
Body weight (kg) = Conversion factor (60 min/hr)
The following equation was used to calculate CADD for A ING3:
Equation 5-10.
Where:
CADD
Duster _wgt
FracTime
Dusting
BW
CF
CADD =
Dustcr wgt x FracTime x Dusting
BW x CF
Potential Chronic Average Daily Dose (mg/kg-day)
Chronic weighted dust concentration (pg/mg)
Fraction of time in environment (unitless)
Dust ingestion rate (mg/day)
Body weight (kg)
Conversion factor (1,000 pg/mg)
The chronic weighted dust concentration was weighted between the dust available from the respirable
portion, floor dust, and abraded particles.
Table 5-13 presents the chronic ingestion doses from each of the underlying models in CEM 3.2 and for
adults and infants 1 to 2 years of age. All lifestages were analyzed. For more information on the
consumer dermal exposure inputs, equations, results (for all lifestages) and sensitivity analysis please
see Appendix J and EPA's CEM 3.0 Appendices (U.S. EPA 2019d).
Table 5-13. Chronic Ingestion Average Daily Doses (mg/kg-day) of TCEP from Consumer Article
Modeling for Adults and Infants 1 to 2 Years of Age (CEM 3.2
COU Subcategory
Consumer Scenario
Lifestage
A ING1
Ingestion
after
Inhalation
AING2
Mouthing
A ING3
Ingestion
of Dust
Total Chronic
Ingestion
ADD
Fabric and textile
products
Carpet back coating
Adult
2.23E-08
0
2.48E-05
1.43E-03
Infant
8.12E-08
2.70E-01
3.15E-04
1.82E-02
Textile - outdoor play
structures
Adult
2.80E-06
0
3.02E-04
1.44E-02
Infant
1.02E-05
2.70E-01
3.84E-03
1.83E-01
Foam auto
Adult
5.78E-10
0
3.22E-10
3.97E-04
Page 192 of 638
-------
COU Subcategory
Consumer Scenario
Lifestage
AING1
Ingestion
after
Inhalation
AING2
Mouthing
AING3
Ingestion
of Dust
Total Chronic
Ingestion
ADD
Foam seating and
bedding product
Infant
2.11E-09
2.70E-01
4.09E-09
5.04E-03
Foam living room
Adult
6.90E-12
0
7.84E-10
7.94E-03
Infant
2.51 E— 11
2.70E-01
9.95E-09
1.01E-01
Mattresses
Adult
3.05E-10
0
1.45E-07
9.66E-04
Infant
1.1IE—09
2.70E-01
1.84E-06
1.23E-02
Other foam objects
Adult
8.81E-12
0
1.05E-09
6.87E-03
Infant
3.21E-11
1.35
1.33E-08
8.72E-02
Building/construction
materials - insulation
Roofing insulation
Adult
6.61E-06
0
7.19E-03
2.12E-02
Infant
2.41E-05
0
9.13E-02
2.70E-01
Acoustic ceiling
Adult
5.01E-07
0
2.44E-04
4.01E-01
Infant
1.83E-06
0
3.10E-03
5.10
Building/construction
materials - wood and
engineered wood
products - wood resin
composites
Wood flooring
Adult
4.24E-06
0
1.46E-03
1.07
Infant
1.55E-05
2.70E-01
1.85E-02
1.36E01
Wood TV stand
Adult
2.78E-08
0
9.53E-06
1.07
Infant
1.01E-07
2.70E-01
1.21E-04
1.36E01
For children and infants, mouthing was the dominant route of exposure. For teenagers and adults,
ingestion of dust was the dominant route of exposure as no mouthing of the consumer articles are
expected.
Sensitivity analyses indicated that "Area of article mouthed" was the driver for the mouthing estimates.
The area of article mouthed was more important for the ingestion estimate compared to the initial
concentration of the SVOC in the article, the density of the article, the surface area of the article, and the
duration of article contact.
For more information on the consumer ingestion exposure inputs, equations, results (for all lifestages)
and sensitivity analysis please see Appendix J and EPA's CEM, Version 3.2 User Guide and
Appendices (U.S. EPA 2023).
5.1.2.2.5 Qualitative Exposure Assessment
Paints and Coatings
A review of literature reporting TCEP used outside the US from the early 2000s provides some evidence
of the use of TCEP in paints and coatings. Ingerowski et al. (2001) detected TCEP in 85 percent of 983
household products in Germany and reported TCEP in wood preservation coatings at a concentration of
10,000 mg/kg (1.0%). Haumann and Thumulla (2002) detected TCEP in paints at a maximum of 840
mg/kg (0.084%) in Germany prior to 2002 (TERA. 2013).
Table 5-14 is a summary of the information gathered for the commercial use of paints and coatings
COU. This data indicate TCEP is used at a high-end of 25 percent in commercial paints and coatings.
Page 193 of 638
-------
Table 5-14. Summary of Commercial Paints and Coatings Concentrations and Density of TCEP
Paint Products
TCEP Concentration
(Mass Fraction)
Product Density (kg/m3)
Low-End
High-End
Low-End
High-End
7 Industrial and commercial paints and
coatings
0.1%
25%
1,000
1,490
Consumer exposures to articles (including automotive articles and replacements parts) that have been
coated with TCEP-containing paints and coatings will mimic consumer exposures from the article
scenarios (e.g., acoustic ceilings, wood resin products). The paints and coatings scenario within CEM
3.2 is for the active application of paints and coatings in a product scenario. Thus, for this risk
evaluation, the dried paints and coatings scenario can be considered a part of the quantitatively assessed
articles scenarios.
The maximum weight fractions (25%) presented in Table 5-14. are up to 4 times higher than the weight
fractions available for consumer articles (6.8%). This suggests that commercial and industrial products
contain higher levels of TCEP than products and articles available for the consumer market. Although it
is possible for consumers to obtain commercial paints and coatings for applications, it is not probable,
and EPA has determined that it is not likely for consumers to obtain TCEP containing paints and
coatings products that are available for commercial applications.
The dermal route is the most important route to consider for exposures to paints and coatings containing
TCEP. The occupational dermal exposure estimates for workers using TCEP-containing paints and
coatings are presented in Section 5.1.1.3. The commercial use of paints and coatings results in a high-
end exposure estimate of 8.02 mg/day and a central tendency estimate of 1.48 mg/day (see Table 5-6).
This scenario is based on a spray application scenario under working conditions for non-occluded
scenarios.
Differences in the occupational and consumer exposure scenarios of paints and coatings provide context
to this qualitative assessment. Products available for the industrial and commercial market are
formulated differently than for consumers. Moreover, workers work with industrial grade formulations
that have higher concentrations of TCEP and may be exposed to paints and coatings containing TCEP
under exposures scenarios that result in higher exposures (e.g., spray application vs. typical domestic
painting).
Wastewater, Liquid Wastes, and Solid Wastes
At the end of their life cycles, consumer articles may be disposed of in municipal solid waste landfills,
construction, and demolition landfills, or undergo incineration. Groundwater monitoring data in Section
3.3.3.5 suggests that TCEP can migrate from municipal unlined landfills to groundwater via landfill
leachate. Water discharges from laundered clothing that picks up TCEP may also be a potential source
of TCEP to surface waters. The successive sections attempt to describe TCEP exposures that may be a
result of the disposal, demolition and removal of household articles and dust containing TCEP. Due to
the difficulties in source attribution, EPA was unable to relate consumer COUs to these TCEP
exposures. However, they are qualitatively discussed to capture additional ways individuals may be
exposed to TCEP via consumer articles.
Page 194 of 638
-------
Wastewater: Section 3.3.2.7 states that laundry wastewater may contribute to elevated environmental
surface water concentrations of TCEP. Clothing has been hypothesized to act as a sink for TCEP to
transfer organophosphate esters from the indoor environment to surface waters via wastewater from
domestic and commercial laundry sources (Schreder and La Guardia. 2014). A study investigating the
relationship between the fate of phthalates and flame retardants transferring from clothing to laundry
wastewater found that chemicals with a log Kow less than 4 showed a greater than 80 percent release to
laundry water, whereas chemicals with a log Kow greater than 6 only showed less than 10 percent
release to laundry wastewater (Saini et al.. 2016). Furthermore, these findings also suggest that dermal
exposure to TCEP may be enhanced from clothing to sweat (Saini et al.. 2016).
TCEP was among the 10 most frequently found compounds, detected at 61.9 percent in wastewater
samples (maximum of 0.7 |ig/L), in a study that collected wastewater from multiple sites in Research
Triangle Park area of North Carolina between 2002 and 2005 (Giorgino et al.. 2007). Flame retardants
were measured primarily at sites downstream from municipal wastewater discharges and at a site
downstream from an industrial fire. TCEP samples were detected in four of eight sites, and at three of
three sites that had major upstream wastewater discharges. A possible explanation for TCEP detection at
the one other site (without an upstream wastewater discharge) was that a fire at an industrial cleaning-
supply warehouse occurred upstream a few months before the sampling event. It is believed that water
applied to control the fire had entered the nearby tributary. In addition, two of these sites near
wastewater discharges are also located near state recreation areas where public facilities, campgrounds,
dump stations, swimming beaches and boating access are available (Giorgino et al.. 2007).
Solid Wastes: A CDC NIOSH report evaluated the occupational exposure to flame retardants at four
gymnastics studios in the mid-2010s. The researchers sampled old foam blocks, mats, padded equipment
and employees via hand wipe samples before and after work. TCEP was detected at 343 ng/ft2 at one of
the gymnastics studios in June 2014, but was not detected in April 2015 after the replacement of new
foam blocks (Broadwater et al.. 2017). A similar study measured 1.6 to 1.9 |ig/g dry weight of TCEP in
polyurethane foam blocks in a Seattle gym. TCEP was detected at a mean concentration of 1.18 |ig/g dry
weight in gym dust concentrations across four gyms. Dust samples were collected from the homes of
four gym instructors. TCEP was found at a mean concentration of 2.5 |ig/g dry weight at the instructors'
residences (La Guardia and Hale. 2015).
A study from the Sierra Nevada foothills suggests that the presence of TCEP on the surfaces of
ponderosa pine needles can be explained by the aerial transport and deposition from nearby point
sources where chemicals were released during the incineration of plastic waste articles (Aston et al..
1996).
Recycling: TCEP is not typically used in electronics (Stapleton et al.. 2011). A CDC NIOSH report
assessed employee exposure to flame retardants at an electronics recycler in November 2016 and
February 2017. TCEP was detected in surface wipe samples at the disassembly workstation at 154
ng/100 cm2. The report indicated the workers were incorrectly wearing N95 respirators and were dry
sweeping. To prevent exposure to airborne TCEP dust particles, the report recommends prohibiting dry
sweeping to clean work areas (Grimes et al.. 2019).
Landfills: The demolition and removal of consumer articles may result in exposures to TCEP.
Construction waste and old consumer products can be disposed of in municipal solid waste landfills and
construction and demolition landfills. Section 3.3.3.8 models the resulting groundwater concentration
that may occur from leaching of TCEP from landfills. Section 3.3.3.63.3.3.8 highlights suspected
leaching of TCEP from nearby landfills (Norman Landfill, Himco Dump, and Fort Devens) (Buszka et
Page 195 of 638
-------
al.. 2009; Barnes et al.. 2004; Hutchins et al.. 1984). The Himco Dump is a closed unlicensed landfill
that included a 4-acre construction debris area. EPA issued a notice in the Federal Register finalizing the
deletion of part of the Himco Dump Superfund site from the National Priorities List (NPL). The Indiana
Department of Environmental Management (IDEM) formally concurred with EPA's proposal on
January 26, 2022, and EPA proposed the site for partial deletion in March, 2022. Fort Devens is also an
EPA superfund site, a former army instillation site that was established in 1917 and closed in 1996, is
also a closed superfund sites. TCEP was detected throughout the entire length of a leachate plume near a
municipal landfill (subtitle D) near Norman, Oklahoma (Barnes et al.. 2004). Leachate samples from
landfill sites in Japan detected TCEP at ranges from 4.1 to 5430 mg/mL. This study suggested that the
origin may be due to plastic wastes (Yasuhara. 1995).
Moran et al. (2023) described that open-air landfills and local waste disposal practices, rather than
former defense sites, may be an important source of atmospheric TCEP in the Arctic. A study around
Troutman Lake, AK indicated higher deposition values on the north side of the lake, which
corresponded with locations of a landfill (900 m from the NW sampling site and 1,400 m from the NE
sampling site). The north-west side of Troutman Lake had the highest deposition with a magnitude of
1,300 ng/m2/day. Troutman lake lies directly south of the village of Gambell on the NW corner of
Sivuqaq. The island is home to the Sivuqaq Yupik people who practice a traditional subsistence
lifestyle. Although the lake was a former chemical disposal site used by the military, Moran et al. (2023)
suggests that the military site was closed (1950) prior to the use of TCEP.
Without a full characterization of non-hazardous landfill (e.g., Norman Landfill) conditions and
historical wastes (e.g., Himco dump and Ft. Devens) around the country, EPA is uncertain how often
contaminant migration occurs given modern practices of non-hazardous landfill and historical site
management. However, the possibility of exposure to TCEP after the release from disposal of consumer
wastes exists.
5.1.2.3 Summary of Consumer Exposure Assessment
Page 196 of 638
-------
Table 5-15. Summary of Acute Daily Rate of Consumer Artie
es Modeled with CEM 3.2
COU Subcategory
Consumer Exposure
Scenario
Lifestage
Exposure Dose (mg/kg/day)
Oral
Inhalation
Dermal
Fabric and textile products
Carpet back coating
Adult
2.43E-04
5.12E-02
6.56E-04
Children
4.00E-02
N/A
1.05E-03
Textile for children's
outdoor play structures
Adult
3.84E-03
1.06
2.33E-02
Children
9.12E-02
N/A
3.73E-02
Foam seating and bedding product
Foam automobile
Adult
2.82E-07
2.90E-04
5.65E-03
Children
3.66E-02
N/A
9.39E-03
Foam living room
Adult
1.86E-07
5.19E-04
1.26E-02
Children
3.66E-02
N/A
2.10E-02
Mattress
Adult
3.50E-06
3.15E-03
1.55E-03
Children
3.66E-02
N/A
2.57E-03
Foam - other (toy block)
Adult
2.47E-07
7.02E-04
8.96E-04
Children
1.83E-01
N/A
1.49E-03
Building/construction materials -
insulation
Roofing insulation
Adult
8.87E-02
2.32E01
1.29E-02
Children
1.27
N/A
2.07E-02
Acoustic ceiling
Adult
5.91E-03
5.31
1.90E-03
Children
8.45E-02
N/A
3.05E-03
Building/construction materials -
wood and engineered wood products
wood resin composites
Wood flooring
Adult
1.07E-01
1.80E02
5.42E-01
Children
1.57
N/A
8.71E-01
Wooden TV stand
Adult
7.03E-04
1.18
7.88E-02
Children
4.66E-02
N/A
1.27E-01
Page 197 of 638
-------
Table 5-16. Summary of Chronic Average Daily Doses of Consumer Articles IV
odeled with CEM 3.2
COU Subcategory
Consumer Exposure
Scenario
Lifestage
Exposure Dose (mg/kg/day)
Oral
Inhalation
Dermal
Fabric and textile products
Carpet back coating
Adult
2.48E-05
4.67E-03
3.50E-04
Children
3.69E-02
N/A
5.63E-04
Textile for outdoor
children s outdoor play
structures
Adult
3.05E-04
6.05E-02
1.35E-02
Children
4.09E-02
N/A
2.17E-02
Foam Seating and Bedding Product
Foam automobile
Adult
9.01E-10
7.95E-07
5.65E-03
Children
3.66E-02
N/A
9.38E-03
Foam living room
Adult
7.90E-10
1.42E-06
1.26E-02
Children
3.66E-02
N/A
2.10E-02
Mattress
Adult
1.45E-07
6.79E-05
1.54E-03
Children
3.66E-02
N/A
2.55E-03
Foam - other (toy block)
Adult
1.05E-09
1.93E-06
8.69E-04
Children
1.83E-01
N/A
1.44E-03
Building/construction materials -
insulation
Roofing insulation
Adult
7.20E-03
1.42
1.03E-03
Children
1.03E-01
N/A
1.66E-03
Acoustic ceiling
Adult
2.45E-04
1.15E-01
7.36E-05
Children
3.50E-03
N/A
1.18E-04
Building/construction materials -
wood and engineered wood products
wood resin composites
Wood flooring
Adult
1.46E-03
1.01
2.41E-01
Children
5.75E-02
N/A
3.87E-01
Wooden TV stand
Adult
9.56E-06
6.57E-03
7.69E-02
Children
3.67E-02
N/A
1.23E-01
Page 198 of 638
-------
Table 5-17. Summary of Lifetime Average Daily Doses of Consumer Articles Modeled with CEM
3.2
COU Subcategory
Consumer Exposure
Scenario
Exposure Dose (mg/
kg/day)
Oral
Inhalation
Dermal
Fabric and textile products
Carpet back coating
1.63E-02
6.04E-03
3.77E-04
Textile for outdoor
children's outdoor play
structures
1.70E-02
7.83E-02
1.45E-02
Foam seating and bedding
product
Foam automobile
1.63E-02
1.03E-06
6.19E-03
Foam living room
1.63E-02
1.84E-06
1.38E-02
Mattress
1.63E-02
8.78E-05
1.68E-03
Foam - other (toy block)
8.14E-02
2.49E-06
9.53E-04
Building/construction materials -
insulation
Roofing insulation
5.26E-03
1.04
7.55E-04
Acoustic ceiling
5.83E-04
1.48E-01
7.92E-05
Building/construction materials -
wood and engineered wood
products - wood resin composites
Wood flooring
1.98E-02
1.30
2.59E-01
Wooden TV stand
1.63E-02
8.50E-03
8.27E-02
5.1.2.4 Weight of Scientific Evidence Confidence for Consumer Exposure
The overall exposure confidence for the various consumer scenarios ranged from slight to moderate.
Slight confidence in the exposure estimates were mainly due to data uncertainties. Information on article
weight fraction was sparse, and it was unclear whether many of the literature values were still relevant
for articles used today. EPA considered a worst-case approach to consumer weight fraction and varied
this parameter in the sensitivity analysis as reported in Appendix J (Consumer Exposure). Information
on exposure scenarios (e.g., mouthing durations, use durations, frequency of contacts per day) were also
limited. Furthermore, limited monitoring data were available to corroborate the modeled consumer
exposure estimates and validate current use of TCEP in consumer articles. In addition, there are
uncertainties related to CEM 3.2 modeling approaches (e.g., deterministic vs. stochastic approaches,
background concentrations, assumptions for dermal absorption parameters).
Page 199 of 638
-------
Table 5-18. Weight of Scientific Evidence Confidence for Chronic Consumer Exposure Modeling Scenarios
Consumer COU
Confidence
in Model
Used"
Confidence
in Model
Default
Values6
Confidence in User-Selected Varied Inputsc
Monitoring
Data
Overall
Exposure
Confidence'
Category
Subcategory
Form
Density
Usedrf
Use
Duration2
Weight
Fraction^
Room of
Use®
Dermal
Kp9 Fabs}
Mouthing'1
Fabric and
textile
products
Carpet back
coating
Article
++
+++
++
+++
++
+++
+
Limited
Moderate
Textile for
outdoor
children's
outdoor play
structures
Article
+++
+
++
++
++
++
++
Limited
Moderate
Building/
construction
materials -
insulation
Roofing
insulation
Article
++
++
+
N/A
+
+++
+
None
Slight
Acoustic
ceiling
Article
+
++
+
N/A
+
++
+
Limited
Slight
Foam
seating and
bedding
product
Foam
automobile
Article
+++
+++
++
++
++
+++
+
Limited
Moderate
Foam living
room
Article
+++
+++
++
+++
++
+++
++
Limited
Moderate
Mattress
Article
+++
+++
++
+++
+
+++
+
None
Slight
Foam - other
(toy block)
Article
+++
+++
++
++
+
+++
++
None
Slight
Building/
construction
materials -
wood and
engineered
wood
products -
wood resin
composites
Wood flooring
Article
+++
+++
++
+++
+
+++
+
None
Slight
Wooden TV
stand
Article
+++
+++
++
++
+
+++
+
Limited
Moderate
Page 200 of 638
-------
Consumer COU
Confidence
in Model
Used"
Confidence
in Model
Default
Values6
Confidence in User-Selected Varied Inputsc
Monitoring
Data
Overall
Exposure
Confidence'
Category
Subcategory
Form
Density
Usedrf
Use
Duration2
Weight
Fraction^
Room of
Use®
Dermal
Kp9 Fabs}
Mouthing'1
11 Confidence in Model Used considers whether model has been peer reviewed, as well as whether it is being applied in a manner appropriate to its design and
objective. The model used (CEM 3.2) has been peer reviewed, is publicly available, and has been applied in a manner intended, to exposures associated with
uses of household products and/or articles. Medium was selected for the carpet-back coating scenario and a roofing insulation scenario because of uncertainties
surrounding the barrier layers. Low was selected for acoustic ceiling because the related CEM scenario was Drywall, and these products have different product
characteristics.
h Confidence in Model Default Values considers default value data source(s) such as building and room volumes, interzonal ventilation rates, and air exchange
rates. These default values are all central tendency values (i.e., mean or median values) sourced from EPA's Exposure Factors Handbook (U.S. EPA. 2017d.
2011a). Low was selected for outdoor play structures, as there were uncertainties on the area volumes related to this scenario.
c Confidence in User-Selected Varied Inputs considers the quality of their data sources, as well as relevance of the inputs for the selected consumer condition of
use.
d Density Used was primarily available for product descriptions. (Westat. 1987)
'' Use Duration is primarily sourced from the EPA's Exposure Factors Flandbook and by the judgment of the exposure assessor.
' Weight fraction of TCEP in articles was sourced from the available literature and database values.
g Room of use (zone 1 in modeling) is informed by professional judgment of the exposure assessor based on the article scenario. The reasonableness of these
judgments is considered in the reported confidence ratings.
h The dermal permeability coefficient (Kp) used (0.022 cm/hr) and fraction absorbed (Fabs) used (35.1%) was derived from a study of TCEP tested on human ex
vivo skin (Abdallah et al.. 2016). Frequency of mouthing (Low, Medium, High) was estimated using the assessors" judgment when considering the exposure
scenario. Literature values override (Poet et al.. 2000) CEM 3.2 default values for fraction absorbed.
' + + + Robust confidence suggests thorough understanding of the scientific evidence and uncertainties. The supporting weight of scientific evidence outweighs
the uncertainties to the point where it is unlikely that the uncertainties could have a significant effect on the exposure estimate.
+ + Moderate confidence suggests some understanding of the scientific evidence and uncertainties. The supporting scientific evidence weighed against the
uncertainties is reasonably adequate to characterize exposure estimates.
+ Slight confidence is assigned when the weight of scientific evidence may not be adequate to characterize the scenario, and when the assessor is making the
best scientific assessment possible in the absence of complete information. There are additional uncertainties that may need to be considered.
Page 201 of 638
-------
5.1.2.4.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for
the Consumer Exposure Assessment
EPA recognizes the need to include an uncertainty analysis. One important distinction for such an
analysis is variability vs. uncertainty—both aspects need to be addressed. Variability refers to the
inherent heterogeneity or diversity of data in an assessment. It is a quantitative description of the range
or spread of a set of values and is often expressed through statistical metrics, such as variance or
standard deviation, which reflect the underlying variability of the data. Uncertainty refers to a lack of
data or an incomplete understanding of the context of the risk evaluation decision.
Variability cannot be reduced but can be better characterized. Uncertainty can be reduced by collecting
more or better data. Quantitative methods to address uncertainty include non-probabilistic approaches
such as sensitivity analysis and probabilistic or stochastic methods. Uncertainty can also be addressed
qualitatively by including a discussion of factors such as data gaps and subjective decisions or instances
where professional judgment was used.
Uncertainties associated with approaches and data used in the evaluation of consumer exposures are
described below. A sensitivity analysis was conducted for the following COUs to understand the drivers
for the inhalation, ingestion, and dermal estimates (Table 5-19).
Table 5-19. Sensitivity Analysis for Chronic Consumer Exposure Modeling Scenarios
Consumer COU
User-Selected Varied Inputs"
Subcategory
Consumer
Exposure
Scenario
Initial SVOC
Concentration in
Article (mg/cm3)6
Mouthing
Duration
(min)c
Surface
Area of
Article (m2)
Events
per
day (n)
Results
Fabric and
textile products
Textile for
outdoor
children's play
structures
4.03
0.93
0.30
High
(8.4/7/10)
Low
(2.3/3.65/5)
Mouthing duration is
a driver of ingestion
exposures.
Building/
construction
materials -
insulation
Roofing
insulation
0.594
0.180
0.06
SVOC concentration
is a driver of
inhalation
exposures.
Building/
construction
materials -
wood and
engineered
wood products
- wood resin
composites
Wood flooring
30
12
211
105
10
5
SVOC concentration
is a driver of dermal
exposures.
Surface area of the
article and Events
per day (n) influence
the dermal exposure
estimates
11 User selected inputs were varied for each of the listed consumer exposure scenarios.
h Initial SVOC concentration in article is a function of the product weight fraction and article density.
c The high mouthing duration defaults in CEM 3.0 were 10 min/event for an infant (< 1 year of age), 7 min/event for
an infant aged 1-2 years, and 8.4 min/event for a child 3-5 years. EPA modified the mouthing durations to 5
min/event for infants < 1 years, 3.65 min/event for 1-2 years, and 2.3 min/event for children 3-5 years to test the
sensitivity of this parameter.
-------
A clear finding of the sensitivity analysis indicated that the initial SVOC concentration (a product of the
density and weight fraction) was a significant driver in the inhalation and dermal exposure estimates for
all scenarios. The initial SVOC concentration was also relevant for the ingestion estimate for the
insulation scenario, likely because there was no estimate for direct mouthing of this COU. Mouthing
duration is an important driver of ingestion exposures for children's play structures. For full results on
the sensitivity analysis please refer to Appendix J (Consumer Exposures).
In the absence of parameter information from the literature, EPA used scientific judgement to select
parameters for consumer modeling. There are uncertainties associated with any scientific judgment. The
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File: Consumer
Exposure Modeling Inputs (U.S. EPA. 2024e) provides a full list of parameters and description of
rationale as to why certain parameter values were selected.
Weight Fraction
The key uncertainty in the consumer exposures assessment was the availability of relevant article weight
fractions data. The Ecology Washington database was the main source of weight fraction information
for the fabric, textile, and leather products scenarios. The 1.3 percent weight fraction for textiles in
outdoor play structures was based on a value from the Washington State Database where the maximum
weight fraction of 67 articles was 1.3 percent (WSDE. 2023). Of the 67 articles, there were only 2 that
contained TCEP. The other article had a level of TCEP of 0.5 percent. Additionally, the database
indicated four detects of TCEP in carpet padding and rug mats (ranged from 0.01 to 0.02%). This
illustrates the limited data availability of weight fraction information for the fabric and textile products
scenario.
The building and construction products scenario (e.g., insulation, acoustic ceiling, wood resin products)
relied on old, foreign literature values from Ingerowski et al. (2001) as cited in SCHER (2012).
Anecdotal information from the literature suggested TCEP is present in these products but did not have
specific information on weight fraction and article concentrations.
Values from Fang et al. (2013) were used to estimate weight fractions for foam seating and bedding
products. There are uncertainties in these estimates because concentrations of V6 (a dimer of TCEP)
were utilized in determining a TCEP weight fraction. This study measured TCEP at 14 percent as an
impurity in V6, and hence this proportion was used to estimate weight fractions of foam seating and
bedding products (Fang et al.. 2013). There are uncertainties associated with how much TCEP is present
as an impurity in V6.
TCEP in articles are not captured in CDR or Datamyne databases, as Datamyne does not include
articles/products containing the chemical unless the chemical name is included in the description. Based
on descriptions provided on the bills of lading, Figure 1-3 provides an estimate of the volume of TCEP
imported as the chemical (not in an identified product or article) from 2012 to 2020. This limitation
further illustrates the difficulty in obtaining current concentrations and weight fractions of TCEP in
consumer products.
Duration and Frequency of Contact and Mouthing
For the carpet back coating scenario and wood flooring scenario, a literature value indicated that
children under 12 years old spend 19 hours per day indoors (EFH 2011). It was assumed that the
frequency of contact per day was 5 events for carpet and 10 events for flooring, and that the area
mouthed was 10 cm2. It should be noted that these values are conservative assumptions for duration and
Page 203 of 638
-------
frequency of contact {i.e., typical frequency may be less than these estimates). The dermal exposure
estimates are sensitive to the frequency of events per day parameter.
A further limitation for the carpet back coating and insulation scenario is the presence of a boundary
layer {e.g., top of the carpet, drywall in between insulation and living space) between the TCEP
containing material and the potentially exposed human {e.g., infant, child, adult). CEM 3.2 uses an
overall mass transfer coefficient that is empirically estimated from an equation based on the AMEM
guidance (the complexity of individual phase mass transfer is subsumed into an overall mass transfer
coefficient that is either measured or estimated from a regression equation based on assorted chemical
measurements). Although CEM 3.2 does not explicitly consider a boundary layer in its modeling, this
does not mean that the model does not attempt to capture this complexity. Nevertheless, it is an
uncertainty associated with the consumer modeling for the scenarios where a boundary layer would be
expected. The modeling as conducted suggests that the TCEP would migrate to the surface of the carpet
from the back coating components, or the dust particles would migrate from the insulation behind the
drywall to the living area.
Oral ingestion estimates are driven by mouthing of articles for infants and children. A sensitive
parameter driving these estimates is the duration of mouthing parameters. The recommended estimates
from CEM 3.2 are 8.4 min/hr, 7 min/hr, and 10 min/hr for young children (aged 3-5 years), infants (1-2
years), and infants (<1 year), respectively.
Trends and Monitoring Data
The paucity of monitoring information related to the consumer COUs makes it difficult for EPA to have
confidence in whether the consumer articles are nationally representative. Moreover, the decreasing
trend of TCEP use, seen in the production volume data and environmental monitoring data, coupled with
the understanding that many manufactures have replaced TCEP with alternatives in their products, build
more uncertainty about the relevance of the consumer modeling to current consumers.
A systematic review revealed that there is limited information related to weight fractions of TCEP in
consumer articles. No SDS were available for TCEP in consumer products. For the limited monitoring
and experimental literature that was available, it is unclear how relevant the concentrations of TCEP at
the time of sampling is related to consumer articles that are produced today.
In 2013, the State of California amended Technical Bulletin 117, a residential upholstered furniture
flammability standard that was first implemented in 1975. The original TB 117 required interior filling
materials of upholstered furniture to withstand exposure to a 12 second small open flame (the small
flame impingement test, a one second flame, and the open flame test). This was replaced with a smolder
resistance test, which tests a lighted cigarette on the fabric outside of the foam in 2013. TB 117-2013 is
of significance to consumer articles, particularly fabric and textiles, and foam seating and bedding
products, as article manufacturers no longer are required to meet the stringent flame standards of TB
117. Flame retardant concentrations in these articles are expected to decrease following this change. The
available monitoring and experimental data on TCEP used in this consumer assessment was gathered
pre-2013 (Table 5-20).
Page 204 of 638
-------
Table 5-20. Summary of Sampling Date for TCEP Weight Fraction Data
COU Subcategory
Weight Fraction Selected
Source
Sampling Date
Fabric and textile
products
• 0.02% carpet back coating
• 1.3% fabric in children's play
structures
Ecology Washington
database WSDE (2023)
2012
Foam seating and
bedding products
• 0.51% furniture foam
• 0.74% auto foam
• 0.64% toy foam blocks
Fans et al. (2013)
2009-2011
Building/construction
materials - insulation
• 1.98% insulation
• 6.8% acoustic ceiling
Inaerowski et al. (2001)
<2001
Building/construction
materials - wood and
engineered wood
products - wood
resin composites
• 3% hardwood floors, wooden
TV stand
SCHER (2012)
1997"
11 Ionas et al. (2014) did provide more recent (2007) data on TCEP in wood tovs at 0.0004%. However, due to the
recent evidence suggesting TCEP use in wooden TV stands, and because TB 117-2013 is relevant for upholstered
foam and furniture materials, EPA selected a weight fraction of 3% for consumer modeling.
Due to the limited information available on article weight fractions, EPA was unable to select a range of
weight fraction for each of the COUs, and rather proceeded to assess consumer exposures to TCEP
containing articles with a single discrete weight fraction value per article scenario. Additional sensitivity
analysis varying the initial SVOC concentration in the article was conducted to help characterize the
results (Table 5-19).
Ionas et al. (2014) stratified their data on TCEP in toys by time of manufacture (before and after 2007
when the REACH regulation went into force). Pre-2007, TCEP was detected in 32 percent of 63
children's toys whereas post-2007 TCEP was detected in 22 percent of 51 children's toys. Nevertheless,
consumer modeling was conducted with possible weight fractions to understand the potential exposure
of such products in furnishings and the built consumer environment.
Table 5-21 summarizes the indoor air and indoor dust monitoring data that was available in the United
States. For a description of statistical methods, methodology of data integration, and treatment of non-
detects and outliers used to generate these estimates, please see the Supplemental Information File:
Environmental Monitoring Concentrations Reported by Media Type (U.S. EPA. 2024i).
Page 205 of 638
-------
Table 5-21. Summary of Indoor Monitoring Data of TCEP from U.S. Studies
Matrices
Location
Type
Count of Estimates
from Studies
Containing U.S. Data
Unit
Fraction
Average of
Arithmetic
Estimates
Average of
90th Percentile
Estimates
Indoor Air
Public spaces
1
ng/m3
Particulate
2.0
4.6
Residential
1
ng/m3
Vapor/gas
9.5
2.1E01
Indoor Dust
Public spaces
1
ng/g
Dry
8.2E02
1.9E03
Residential
9
ng/g
Dry
1.1E03
2.2E03
Vehicles
1
ng/g
Dry
4.2E03
8.9E03
The maximum SVOC air concentration of 9.32 mg/m3 for the insulation condition of use is five orders
of magnitude higher than the 90th percentile estimate of indoor residential air concentrations found in
one U.S. study (2,1 / 10 5 mg/m3) (Dodson et al.. 2017). The maximum respirable portion dust
concentration of 11.13 |ig/mg (1.1 x 107 ng/g) is four orders of magnitude higher than the 90th percentile
estimate of residential indoor dust concentrations among nine U.S. studies (2.2><103 ng/g).
Modeling Approach Uncertainties
CEM 3.2 is a deterministic model where the outputs are fully determined by the choices of parameter
values and initial conditions. Stochastic approaches feature inherent randomness, such that a given set of
parameter values and initial conditions can lead to an ensemble of different model outputs. The overall
approach to the CEM modeling is intended to capture a range of low- to high-intensity user exposure
estimates by varying only a limited number of key parameters that represent the range of consumer
product and use patterns for each scenario. A limited set of parameters were varied in the sensitivity
analysis described in Table 5-19. Because not all parameters were varied, there is uncertainty regarding
the full range of possible exposure estimates. Although these estimates are thought to reflect the range of
exposure estimates for the suite of possible exposures based on the varied parameters, the scenarios
presented are not considered bounding or "worst-case," as there are unvaried parameters that are also
identified as sensitive inputs held constant at a central tendency value. Because EPA's largely
deterministic approach involves choices regarding highly influential factors such as weight fraction and
mouthing duration, it likely captures the range of potential exposure levels although it does not
necessarily enable characterization of the full probabilistic distribution of all possible outcomes.
CEM 3.2 has a set of predefined consumer exposure scenarios that do not always line up with the
conditions of use. For example, the CEM scenario utilized for consumer exposure to carpet back coating
was Fabrics: curtains, rugs, wall coverings. There are uncertainties on how TCEP migrates from carpet
back coatings to the surface of carpets and rugs. The literature describes that triphosphate esters such as
TCEP have "blooming potential," which refers to the ability for the chemical to diffuse from a rubber or
plastic material to the outer surface after curing (SCHER. 2012). Furthermore, the study from Castorina
et al. (2017) has indicated that TCEP levels in dust are significantly associated with the presence of
extremely worn carpets, suggesting that TCEP can be sampled in the dust from carpets and make it to
the surface.
Background levels of TCEP in indoor air and indoor dust are not considered or aggregated in this
assessment; therefore, there is potential for underestimating consumer exposures. Furthermore,
consumer exposures were evaluated on a COU specific basis and are based on the use of a single
consumer article, not multiple articles in the indoor environment.
Page 206 of 638
-------
There are uncertainties regarding the use of the 35.1 percent dermal fraction absorption (Fabs) parameter
for the consumer dermal exposure estimates. This is the 24-hour value for fraction absorbed from
Abdallah et al. (2016). EPA cannot assume that consumers would immediately wash their hands
following contact with consumer articles. Therefore, it was assumed that the dose that deposited on the
skin during exposure to a consumer article would remain on the skin until the skin was eventually
washed. While the duration of interaction with materials that contain TCEP may be shorter than the
duration that was tested in the dermal absorption study (i.e., a 24-hour exposure), EPA decided to use
the 35.1 percent fraction absorption value from Abdallah et al. (2016). due to uncertainties related to
consumer hand-washing behaviors.
5.1.3 General Population Exposures
TCEP - General Population Exposures (Section 5.1.3):
Key Points
EPA evaluated the reasonably available information for the following general population exposures,
the key points of which are summarized below:
• Oral ingestion for tribal fishers had the highest exposure estimates (0.008 to 219 mg/kg-day)
among all routes. The highest tribal exposure estimates were for the formulation of TCEP
containing reactive resin OES.
• The hypothetical scenario of a child playing in mud near a facility releasing TCEP to the
ambient air resulted in the highest dermal exposures at a maximum of 7.97 mg/kg-day for use
of paints and coatings at job sites OES. Estimates for a child conducting activities with soil
(2.12xl0~3 mg/kg-day) and incidental soil ingestion (1.08xl0_1 mg/kg-day) were calculated.
Paints and coatings were the only OES for the children playing in mud scenario with margin
of exposures (MOEs) below the benchmark for non-cancer as described in Section 5.3.2.3.
• The highest inhalation exposure concentrations were for the use of paints and coatings at job
sites OES at a central tendency estimate of 3.36x 10~5 and a 95th percentile of 8.21 x 10~5
|ig/m3.
• Exposure estimates for drinking water non-dilute from surface water (1,46x 10~4 mg/kg-day)
were highest for the formulation of TCEP containing reactive resins OES.
• Children in fenceline communities and subsistence fishers are PESS who may have elevated
exposures to TCEP compared to the rest of the general population due to industrial and
commercial environmental releases.
General population exposures occur when TCEP is released into the environment and the environmental
media is then a pathway for exposure. Section 3.3 provides a summary of the monitoring, database, and
modeled data on concentrations of TCEP in the environment. Figure 5-4 below provides a graphic
representation of where and in which media TCEP is estimated to be found and the corresponding route
of exposure.
Page 207 of 638
-------
Figure Legend
Negligible
» Low/Slow
Moderate
High/Fast/Strong
Very High/Rapid/Strong
Partitioning/T ransportation
~ T ransformation/Degradation
Wastewater Facility
Figure 5-4. Potential Human Exposure Pathways to TCEP for the General Population"
" The diagram presents the media (white text boxes) and routes of exposure (italics for oral, inhalation, or dermal)
for the general population. Sources of drinking water from surface or water pipes is depicted with grey arrows.
This diagram pairs with Figure 2-1 depicting the fate and transport of the subject chemical in the
environment.
5.1.3.1 Approach and Methodology
TCEP is used primarily as an additive flame retardant in a variety of materials. TCEP has been detected
in the indoor and outdoor environment and in human biomonitoring indicating that some amount of
exposure is occurring in some individuals, although exposures likely vary across the general population.
See Section 3.3 and Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review
Supplemental File: Data Extraction Information for General Population, Consumer, and Environmental
Exposure (U.S. EPA. 2024r) for a summary of environmental and biomonitoring studies where TCEP
has been detected.
Page 208 of 638
-------
Releases of TCEP are likely to occur through the following mechanisms: diffusion from sources, gas-
phase, and particle-phase mass-transfer, abrasion of materials to form small particulates through routine
use, and direct transfer from articles to dust adhered to the article surface. Releases of flame retardants
to the outdoor environment may occur through direct releases to water, land, and air as well as indirect
releases from the indoor environment.
For a more detailed discussion about indoor SVOC exposure, fate, and transport in the indoor
environment, please see Section 2.2.2.
Exposure to the general population was estimated for the industrial and commercial releases per OES.
Table 3-3 illustrates how the industrial and commercial releases to the environmental media varies by
OES.
Modeled air concentrations (see Section 3.3.1.2) were utilized to estimate inhalation exposures (see
Section 5.1.3.2) to the general population at various distances from a hypothetical facility. Modeled
surface water concentrations (see Section 3.3.2.5) were utilized to estimate oral drinking water
exposures, oral fish ingestions exposures, incidental oral exposures (see Section 5.1.3.4), and incidental
dermal exposures (see Section 5.1.3.3) for the general population. Modeled groundwater concentrations
(see Section 3.3.3.8), were also used to estimate oral drinking water exposures (see Section 5.1.3.4) to
the general population. Modeled soil concentrations (see Section 3.3.3.2) via deposition were used to
estimate dermal and oral exposures (see Sections 5.1.3.3 and 5.1.3.4) to children who play in mud and
other activities with soil.
Exposures estimates from industrial and commercial releases of TCEP were compared to exposure
estimates from non-scenario specific monitoring data to ground truth the results (e.g., indoor dust
exposures). Table 5-22 summarizes the environmental media monitoring data that was available in the
United States. For a description of statistical methods, methodology of data integration and treatment of
non-detects and outliers used to generate these estimates please see the Risk Evaluation for Tris(2-
chloroethyl) Phosphate (TCEP) - Supplemental Information File: Environmental Monitoring
Concentrations Reported by Media Type (U.S. EPA. 2024i).
Table 5-22. Summary of Environmental Monitoring Data of TCEP from the Literature for U.S.
Studies
Matrices
Location Type
Count of Estimates
from Studies
Containing U.S. Data
Unit
Fraction
Average of
Arithmetic
Estimates
Average of
90th Percentile
Estimates
Environmental media
Ambient Air
General Population
6
ng/m3
Any
1.3E-01
2.5E-01
Drinking Water
General Population
1
ng/L
Any
4.9
9.5
Sediment
General Population
1
ng/g
Dry
2.3
4.1
Surface Water
General Population
5
ng/L
Any
1.3E02
2.5E02
Wastewater
Treated Effluent
2
ng/g
Wet
2.1E01
4.3E01
Treated Effluent
4
ng/L
Wet
8.1E02
1.2E03
Ecological media
Page 209 of 638
-------
Matrices
Location Type
Count of Estimates
from Studies
Containing U.S. Data
Unit
Fraction
Average of
Arithmetic
Estimates
Average of
90th Percentile
Estimates
Aquatic Fish
General Population
1
ng/g
Lipid
1.0E01
2.5E01
Terrestrial
Birds
General Population
2
ng/g
Wet
5.3
9.7
Terrestrial
Plants
Remote
1
ng/g
Wet
1.3E02
2.2E02
Human biomonitoring
Human Hair
General Population
2
ng/g
Dry
2.7E02
4.2E02
Human Nails
General Population
1
ng/g
Dry
6.3E02
1.4E03
Figure 5-5 depicts the direct and indirect methods EPA used to estimate general population exposures.
The direct assessment used environmental release estimates that were related to the industrial and
commercial OES (see Section 3.2). Release estimates were used to model ambient air concentrations
(see Section 3.3.1.2), surface water concentrations (see Section 3.3.2.5), soil concentrations (see Section
3.3.3.2), and groundwater concentrations as a result of landfill leachate (see Section 3.3.3.8). EPA
modeled estimates for the environmental media were used to estimate inhalation, dermal and ingestion
doses for various anticipated scenarios (e.g., children's dermal exposure to soil, fish ingestion for the
general population, drinking water ingestion exposure). Further information on the assessed exposure
scenarios is presented in the individual sections below. In addition, EPA estimated exposure doses using
an indirect estimation method via reverse dosimetry (see Section 5.1.3.5). Furthermore, to help "ground
truth" the results, the reported environmental monitoring and reported modeled data (i.e., TCEP
concentration and doses in dietary sources, dust, soil, ambient air, indoor air, and surface water) were
compared against the exposure estimates calculated from the direct assessment patterns.
Page 210 of 638
-------
Direct
Assessment
Indirect
Assessment
Figure 5-5. Direct and Indirect Exposure Assessment Approaches Used to Estimate General
Population Exposure to TCEP
For each exposure pathway, central tendency and high-end exposures were estimated. EPA's Guidelines
for Human Exposure Assessment defined central tendency exposures as "an estimate of individuals in
the middle of the distribution." It is anticipated that these estimates apply to most individuals in the
United States. High-end exposure estimates are defined as "plausible estimate of individual exposure for
those individuals at the upper end of an exposure distribution, the intent of which is to convey an
estimate of exposure in the upper range of the distribution while avoiding estimates that are beyond the
true distribution." It is anticipated that these estimates apply to some individuals, particularly those who
may live near facilities with elevated concentrations.
Page 211 of 638
-------
5.1.3.1.1 General Population Exposure Scenarios
Figure 5-6 provides an illustration of the exposure scenarios considered for general population exposure.
Ambient Air Exposure Scenarios
The Ambient Air Methodology utilizing AERMOD evaluated exposures to human populations at eight
finite distances (10, 30, 60, 100, 1,000, 2,500, 5,000, and 10,000 m) and one area distance (100 to 1,000
m) from a hypothetical releasing facility for each OES. Human populations for each of the eight finite
distances were placed in a polar grid every 22.5 degrees around the respective distance ring. This results
in a total of 16 modeled exposure points around each finite distance ring for which exposures are
modeled. Figure 5-6 provides a visual depiction of the placement of exposure points around a finite
distance ring. Although the visual depiction only shows exposure point locations around a single finite
distance ring, the same placement occurred for all eight finite distance rings.
Figure 5-6. Modeled Exposure Points for Finite Distance Rings for Ambient Air Modeling
(AERMOD)
Modeled exposure points for the area distance evaluated were placed in a cartesian grid at equal
distances between 200 and 900 m around each releasing facility (or generic facility for alternative
release estimates). Exposure points were placed at 100-meter increments. This results in a total of 456
points for which exposures are modeled. Figure 5-6 provides a visual depiction of the placement of these
exposure points (each dot) around the area distance ring.
Although the ambient air is a minor pathway for TCEP, the general population may be exposed to
ambient air concentrations and air deposition because of TCEP releases. Relevant exposures scenarios
considered in this risk evaluation include ambient air inhalation for populations living nearby releasing
Page 212 of 638
-------
facilities, and ingestion and dermal exposure of soil to children result of ambient air deposition from a
nearby facility.
Soil Exposure Scenarios
Air deposition fluxes from AERMOD were used to estimate soil concentrations at various distances
from the hypothetical facility for each OES (see Section 3.3.3.2). Oral ingestion and dermal absorption
exposure estimates of soil were calculated for children aged 3 to 6 years. Ingestion estimates were
calculated for a central tendency and high intake rate. Dermal absorption estimates were calculated for
two exposure scenarios: a child playing in mud, and a child performing activities with soil.
Water Exposure Scenarios
TCEP is expected to be found predominantly in water or soil. Section 3.3.2.5 provides modeled
estimates of TCEP in surface water due to release of TCEP to water. Section 3.3.2.6 provides model
estimates of TCEP in surface water due to air deposition to surface waters. Section 3.3.3.8 provides
modeled estimates of TCEP in groundwater due to estimated migration from landfill leachate. Each of
these estimates were used to calculate an exposure dose from drinking water for the general population.
Additionally, modeled surface water concentrations (see Section 3.3.2.5) were used to calculate a dermal
exposure estimate from swimming, incidental ingestion estimates from swimming, fish ingestion
exposure.
5.1.3.2 Summary of Inhalation Exposure Assessment
Modeled ambient air concentrations for various distances from a hypothetical facility for each COU are
presented in Section 3.3.1.2. Figure 5-7 below is a graph of the inhalation concentration by distances for
the low production volume (2,500 lb/year) low-end and high-end estimates by the central tendency and
high meteorology data. The x-axis is in log scale of distances in meters and the y-axis is in log scale of
the 50th percentile concentrations in ppm.
Page 213 of 638
-------
-5.0 H
?
Q.
& -7 5
c
o
CD
•g -100
8
o
c
8
©
Low Release (2500 lbs)
Low Release (2500 lbs)
Central Tendency Estimate
High-End Estimate
^
1
1 T- m ****-,
¦
p
£
£
I
o>
5
i.oJ
-10^
1.0
2.0
20
2 5 3.0 1.0 IS
Log [Distance (m)]
COM3 refers to Use in paints and coatings at /ob sites
IND refers to Use of Lab Chemicals
MFG refers to Repackaging of Import Containers
PROC-article refers to Processing into 2-part resin article
PROC-resin refers to Incorporation into paints and coatings - resins/solvent-borne
PROC-waterborna refers to Incorporation into paints and coatings - waterborne coatings
PROC-reactive refers to Formulation of TCEP containing reactive resin
30
Scenario
COM3
IND
MFG
-+- PROC-aftide
PROC-reactive
PROC-resin
PR OC-waterborne
Figure 5-7. General Population Inhalation Concentrations (ppm) by Distance (in) in Log Scale
Table 5-23 below indicates the ambient air concentrations at one distance (100 m) for each of the OES.
For a full set data for all distances please see Appendix 1.3.
Page 214 of 638
-------
Table 5-23. Excerpt of Ambient Air Modeled Concentrations for the 2,500 lb Production Volume,
OESfl
Meteorology
Source
Concentration (ppm) by Percentile
10th
50th
95th
Use in paints and coatings at job
sites
MetCT
FUGU
1.15E-05
3.36E-05
6.45E-05
MetHIGH
FUGU
8.77E-06
3.08E-05
8.21E-05
Use of laboratory chemicals
MetCT
ALL
1.51E-08
2.04E-08
3.33E-08
MetHIGH
ALL
1.16E-08
2.24E-08
3.32E-08
Repackaging of import containers
MetCT
ALL
1.50E-10
3.88E-10
9.12E-10
MetHIGH
ALL
2.34E-10
4.39E-10
1.12E-09
Processing into 2-part resin article
MetCT
ALL
1.48E-08
1.93E-08
2.70E-08
MetHIGH
ALL
9.46E-09
1.96E-08
2.72E-08
Incorporation into paints and
coatings - 2-part reactive coatings
MetCT
ALL
2.60E-11
1.60E-09
1.14E-08
MetHIGH
ALL
3.46E-10
2.29E-09
1.1IE—08
Incorporation into paints and
coatings - 1-part coatings
MetCT
ALL
4.80E-09
1.31E-08
2.87E-08
MetHIGH
ALL
4.00E-09
1.35E-08
3.51E-08
Formulation of TCEP containing
reactive resin
MetCT
ALL
2.72E-11
1.78E-09
1.26E-08
MetHIGH
ALL
3.73E-10
2.52E-09
1.21E-08
11 Table 3-3 provides a crosswalk of industrial and commercial COUs to OESs
5.1.3.3 Summary of Dermal Exposure Assessment
5.1.3.3.1 Incidental Dermal from Swimming
The general population may swim in surface waters (streams and lakes) that are affected by TCEP
contamination. Modeled surface water concentrations from EFAST 2014 were used to estimate acute
doses and average daily doses because of dermal exposure while swimming.
The following equations were used to calculate incidental dermal (swimming) doses for all COUs, for
adults, youth, and children:
Equation 5-11.
ADR =
SWC x Kp x SA x ET x CF1 x CF2
BW
Equation 5-12.
ADD =
SWC x Kp XSA xETxRD XET x CF 1 x CF2
BW x AT x CF3
Where:
ADR = Acute Dose Rate (mg/kg-day)
ADD = Average Daily Dose (mg/kg-day)
SWC = Chemical concentration in water (pg/L)
Page 215 of 638
-------
Kv =
Permeability coefficient (cm/h)
SA
Skin surface area exposed (cm2)
ET
Exposure time (h/day)
RD
Release days (days/year)
ED
Exposure duration (years)
BW =
Body weight (kg)
AT =
Averaging time (years)
CF1 =
Conversion factor (1.0/ 10 3 mg/|ig)
CF2 =
Conversion factor (1,0x 10 3 L/cm3)
CF3 =
Conversion factor (365 days/year)
A summary of inputs utilized for these exposure estimates are provided in Appendix I.
EPA used the dermal permeability coefficient (Kp) (0.022 cm/h) derived by Abdallah et al. (2016) from
their in vitro study that measured TCEP absorption through excised human skin.
Table 5-24. Modeled Incidental Dermal (Swimming) Doses for all COUs for Adults, Youths, and
Children, for the 2,500 lb High-End Release Estimate
OESfl
Surface Water
Concentration
Adult (> 21 years)
Youth (11-15 years)
Child (6-10 years)
30Q5
Cone.
(Hg/L)
Harmonic
Mean
Cone.
(Hg/L)
ADRpot
(mg/kg-
day)
ADD
(mg/kg-
day)
ADRpot
(mg/kg-
day)
ADD
(mg/kg-
day)
ADRpot
(mg/kg-
day)
ADD
(mg/kg-
day)
Repackaging of import
containers
862.129
1,366.528
1.39E-03
6.02E-06
1.06E-03
4.61E-06
6.44E-04
2.80E-06
Incorporation into
paints and coatings -
1-part coatings
3,819.444
5,912.114
6.14E-03
2.61E-05
4.70E-03
2.00E-05
2.85E-03
1.21E-05
Incorporation into
paints and coatings -
2-part reactive
coatings
3,462.800
5,360.066
5.57E-03
2.36E-05
4.27E-03
1.81E-05
2.59E-03
1.10E-05
Use in paints and
coatings at job sites
2,029.305
3,216.574
3.26E-03
1.42E-05
2.50E-03
1.09E-05
1.52E-03
6.58E-06
Formulation of TCEP
containing reactive
resin
4,844.722
6,245.374
7.79E-03
2.75E-05
5.97E-03
2.11E-05
3.62E-03
1.28E-05
Use of laboratory
chemicals
34.555
54.722
5.59E-05
2.41E-07
4.26E-05
1.85E-07
2.58E-05
1.12E-07
" Table 3-3 provides a crosswalk of industrial and commercial COUs to OES.
5.1.3.3.2 Incidental Dermal Intake from Soil
Dermal absorbed doses (DAD) were calculated for TCEP using the following formula:
Page 216 of 638
-------
Equation 5-13.
Where:
AF =
ABSd =
Csoil =
CF =
SAskin =
EV =
BW =
AF =
Modeled soil concentrations were calculated from 95th percentile air deposition (see Section 3.3.3.2) for
100 and 1,000 m. These calculations were conducted for the COM-paints-use scenario (LOW PV -
2,500 lb, HE-95th percentile release).
Soil concentrations of 141.2 ng/g were modeled for the 2,500 lb production volume, high-end release
estimate for the Incorporation into paints and coatings - 1 part coatings OES using BST (see Section
3.3.3.5).
The dermal absorption fraction {ABSd) used was 35.1 percent (Abdallah et al.. 2016). The skin surface
area for the arms (0.106 m2), hands (0.037 m2), legs (0.195 m2) and feet (0.049 m2), and body weight
(18.6 kg) of a 3- to 6-year-old was used from the Exposure Factors Handbook (U.S. EPA. 2017d). EPA
used two different scenarios for the adherence factor of soil to skin: 96 mg/cm2 for a child playing in
mud and 0.467 mg/cm2 for children's activity with soil. With an assumption of one event per day and an
averaging time of 2 days, the dermal exposure estimates for the different scenarios were as follows:
Table 5-25. Modeled Soil Dermal Doses for the Commercial Use of Paints and Coatings COU, for
Children
n A n _ Csoa x CF x AF x ABSd x SAsk[n x EV
~ BW x AT
Adherence factor of soil to skin (mg/cm2-event)
Dermal absorption fraction
Concentration in soil
Conversion Factor
Skin surface area (cm2)
Events per day
Body weight (kg)
Averaging time
OES
Exposure
Distance
Soil Concentration
Dermal Absorbed Dose
Scenario
(m)
(ng/g)
(mg/kg-day)
Activities
100
1.14E04
3.89E-04
Use in paints and
coatings at job
sites17
with soil
1,000
8.65E01
2.95E-06
Playing in
100
1.14E04
7.99E-02
mud
1,000
8.65E01
6.06E-04
Incorporation into
Activities
N/A
1.41E02
4.82E-06
paints and
with soil
coatings - 1-part
Playing in
N/A
1.41E02
9.90E-04
coatings
mud
11 95th percentile estimates
5.1.3.4 Summary of Oral Exposures Assessment
5.1.3.4.1 Drinking Water Exposure
Page 217 of 638
-------
us
General Population (Background)
I Unknown/Not Specified
V Lognonnal Distribution (CT and 90th percentile)
H Non-Dctcct
Mix
NonUS
4253347 - Padhye et al.t 2014 - US
3975066 - Hopple et al.. 2009 - US
3364193 - Kingsbury el al., 2008 - US
3559503 - Focazio el al., 2008 - PR,US
1487184 - Lebel el al., 1987 - CA.US
3455908 - Lcc ct al.. 2016 ¦ KR
5469210 - Valcarcel et al., 2018 - ES
1250860 - Rodil et al.. 2012 - ES
5469582 - Yasuhara. 1994 - JP
¦ w
t7V
ft
IOA-6
l0A-4
0.01 1
Concentration (ng/L)
100
10*4
Figure 5-8. Concentrations of TCEP (ng/L) in Drinking Water from 1982 to 2014
A study of drinking water systems in the United States indicated a maximum of 470 ng/L and a median
of 120 ng/L of TCEP in finished water, and a maximum of 200 ng/L and a median of 140 ng/L in
distributed waters in 6 out of 19 drinking water systems. The drinking water systems collected samples
from 19 drinking water treatment plants (DWTPs) across the United States, representing drinking water
for more than 28 million Americans (Benotti et al.. 2009).
TCEP has been detected in tap water in Korea at a mean of 39.5 and a maximum of 87.4 ng/L as
recently as 2017 (Park et al.. 2018). Because the OPFR concentrations were correlated with the distance
of the pipes (both from the water intake source to the drinking water treatment facility and the drinking
water treatment facility to the sampling site), this study has suggested that a possible source of OPFRs in
tap water were pipes. Pipe materials are known to promote the formation of disinfection by products or
biofilms (Park et al.. 2018).
Drinking Water Intake Estimates via Modeled Surface Water Concentrations
Modeled surface water concentrations (see Sections 3.3.2.5 and 3.3.2.6) were used to estimate drinking
water exposures. A 0 percent drinking water treatment removal efficiency was used for the purposes of
this exposure estimation.
Drinking water intakes were calculated using the following formulae:
Equation 5-14.
( DWT\
SWC x (1 - x IRdw xRD x CF1
AD RpoT —
BW x AT
Equation 5-15.
( DWT\
SWC x (1 - x IRdw x ED XRD x CF1
ADDpot —
BW x AT x CF2
Page 218 of 638
-------
Equation 5-16.
( DWT\
SWC x (1 - ^rr) x IRdw x ED x RD x CF1
LADDpnr = -
P0T BW x AT x CF2
Equation 5-17.
( DWT\
SWC x (^1 - x ED XRD x CF1
LAD CP0T =
POT AT X CF2
Where:
ADRpot
Potential Acute Dose Rate (mg/kg/day)
ADDpor
Potential Average Daily Dose (mg/kg/day)
LADDpor
Potential Lifetime Average Daily Dose (mg/kg/day)
LADCpot
Potential Lifetime Average Daily Concentration in drinking water
(mg/L)
SWC
Surface water concentration (ppb or |ig/L; 30Q5 cone for ADR,
harmonic mean for ADD, LADD, LADC)
DWT
Removal during drinking water treatment (%)
IRdw
Drinking water intake rate (L/day)
RD
Release days (days/yr for ADD, LADD and LADC; 1 day for
ADR)
ED
Exposure duration (years for ADD, LADD and LADC; 1 day for
ADR)
BW
Body weight (kg)
AT
Exposure duration (years for ADD, LADD and LADC; 1 day for
ADR)
CF1
Conversion factor (1,0x 10 3 mg/|ig)
CF2
Conversion factor (365 days/year)
A method was derived to incorporate a dilution factor to estimate TCEP concentrations at drinking water
locations downstream from surface water release points. Because no location information was available
for facilities releasing TCEP, a dilution factor and distances to drinking water intake was estimated for
each relevant SIC code. Table 5-26 provides the 50th quantile distances and 50th quantile dilution
factors for the 30Q5 and harmonic mean flow for the relevant SIC codes.
Page 219 of 638
-------
Table 5-26. 50th Quantile Distances and 30Q5 and Harmonic Mean 50th Quantile Dilution
Factors for Relevant TC
EP SIC
SIC Codes
n
50th Quantile
Distance
(km)
50th Quantile
Dilution
Factor (30Q5)
50th Quantile Dilution
Factor
(Harmonic Mean)
Adhesives, Sealants,
Plastics, Resins,
Rubber Manufacturing
516
113.82
432.36
528.47
Paint Formulation
374
107.03
1,603.6
1,854.89
POTWs - All facilities
567
129.57
1,233.87
1,557.91
30Q5 = The lowest 30-day average flow that occurs (on average) once every 5 years
To calculate the diluted water concentrations the surface water concentrations from E-FAST 2014
modeling were divided by the dilution factors presented in Table 5-26. Table 5-27 presents the diluted
drinking water concentrations for adults for all industrial and commercial COUs.
Table 5-27. Modeled Drinking Water Ingestion Estimates for Diluted Surface Water
Concentrations for Adults for All Industrial and Commercial COUs for the 2,500 lb High-End
Release Estimate
OESfl
Diluted Water Concentration
Adult (> 21 years)
Harmonic Mean
Concentration
(^g/L)
30Q5
Concentration
(^g/L)
ADRpot
(mg/kg-
day)
ADD
(mg/kg-
day)
LADDpot
(mg/kg-
day)
LADCpot
(mg/L)
Repackaging of import
containers
0.553
1.108
4.46E-05
1.67E-08
7.05E-09
6.41E-07
Incorporation into
paints and coatings -
1-part coatings
2.059
3.687
1.48E-04
6.20E-08
2.62E-08
2.39E-06
Incorporation into
paints and coatings -
2-part reactive
coatings
1.867
3.343
1.35E-04
5.62E-08
2.38E-08
2.16E-06
Use in paints and
coatings at job sites
1.303
2.607
1.05E-04
3.92E-08
1.66E-08
1.51E-06
Formulation of TCEP
containing reactive
resin
9.167
14.445
5.81E-04
2.76E-07
1.17E-07
1.06E-05
Use of laboratory
chemicals
0.022
0.044
1.79E-06
6.68E-10
2.83E-10
2.57E-08
11 See Table 3-3 for a crosswalk of industrial and commercial COUs to OESs.
Table 5-28 provides the non-diluted drinking water intake estimates. In this case, it is assumed that the
surface water outfall is located very close (within a few km) to the population. The dilution factor
reduces the acute, chronic, and lifetime exposure estimates by a factor of three.
Page 220 of 638
-------
Table 5-28. Modeled Drinking Water Ingestion Estimates for Surface Water Concentrations for
OESfl
Water Concentration
Adult (> 21 years)
Harmonic
Mean
Concentration
(Hg/L)
30Q5
Concentration
(^g/L)
ADRpot
(mg/kg-
day)
ADD
(mg/kg-
day)
LADDpot
(mg/kg-
day)
LADCpot
(mg/L)
Repackaging of
import containers
862.129
1,366.528
5.4992E-02
2.60E-05
1.10E-05
9.99E-04
Incorporation into
paints and coatings -
1-part coatings
3,819.444
5,912.114
2.3792E-01
1.15E-04
4.87E-05
4.43E-03
Incorporation into
paints and coatings -
2-part reactive
coatings
3,462.800
5,360.066
2.1570E-01
1.04E-04
4.41E-05
4.01E-03
Use in paints and
coatings at job sites
2,029.305
3,216.574
1.2944E-01
6.11E-05
2.59E-05
2.35E-03
Formulation of TCEP
containing reactive
resin
4,844.722
6,245.374
2.5133E-01
1.46E-04
6.17E-05
5.62E-03
Use of laboratory
chemicals
34.555
54.772
2.20E-03
1.04E-06
4.40E-07
4.01E-05
11 Table 3-3 provides a crosswalk of industrial and commercial COUs to OES.
A summary of inputs utilized for these exposure estimates is presented in Appendix I.2Appendix I.
Drinking Water via Leaching of Landfills to Groundwater
Groundwater concentrations from leaching from landfills was estimated for the 2,500 and 25,000 lb
production volume scenarios (see Table 3-8. in Section 3.3.3.8). The relevant COU/OES that may be
relevant for groundwater migration from landfill leachate are the Incorporation into paints and coatings
- 1-part coatings, and Processing into formulation of TCEP-containing reactive resin. These OESs result
in the following releases to landfill presented in Table 5-29. In addition, consumer articles could be
disposed to municipal solid waste landfills and construction and demolition landfills.
Table 5-29. Landfill Releases of TCEP from Two Commercial and Industrial OESs
OES
Number of Release
Days
Annual Release Per Site
(kg-site-yr)
Daily Release
(kg/site-day)
Incorporation into paints and
coatings - 1-part coatings
2
2.15E01
9.27
Formulation of TCEP-
containing reactive resin
17
4.29E01
2.49
Section 3.3.3.8 estimates a range of groundwater concentrations because of industrial and commercial
releases. The range of concentrations varies due to leachate concentrations to be between 1.08x10 3 and
1.08X101 |ig/L. Using the same formulae for drinking water ingestion above, adult drinking water
estimates because of landfill leachate contamination are presented in Table 5-30.
Page 221 of 638
-------
Table 5-30. Estimated Average Daily Doses and Lifetime Average Daily Concentrations for Adults
from Groundwater Concentrations by DRAS
DRAS
Groundwater
Concentration
Adult (>21 years)
ADD
(mg/kg-day)
LADCpot
(mg/L)
Low Estimate: Low Leachate Concentration -
2,500 lb Production Volume
1.08E-03
3.3E-11
3.0E-09
High Estimate: High Leachate Concentration -
2,500 lb Production Volume
1.08E01
3.3E-07
3.0E-05
These results would be further lowered if dilution was incorporated to these drinking water estimates.
Due to uncertainties in distance from drinking water intake location to the groundwater contamination
site the dilution was not estimated.
The complete set of drinking water exposure estimates can be referenced in Supplemental Information
File: E-FASTModeling Results (U.S. EPA. 2024s).
5.1.3.4.2 Fish Ingestion Exposure
Surface water concentrations for TCEP associated with a particular COU were modeled using E-FAST
2014 as described in Section 3.3.2.5. Surface water concentrations based on harmonic mean surface
water flows representing the 50th percentile stream flows of all facilities in each industry sector were
used to estimate exposure from fish ingestion. The harmonic mean flow was used to estimate the
concentration of TCEP in fish tissue because it represents long-term average flow conditions. As it takes
time for chemical concentrations to accumulate in fish, a harmonic mean flow is more appropriate than a
low streamflow value (e.g., 7Q10) that occurs infrequently. Furthermore, dilutions of surface water
concentrations of TCEP further downstream of a facility's outfall was not considered, as fish
presumably reside within stream reaches receiving direct releases from a facility. This approach takes
into account that people often harvest fishes originating from various locations regardless of known or
unknown releases to the environment at that location; thus, it is more conservative because it estimates
higher concentrations of TCEP in fish.
General population exposure estimates from fish consumption are provided for only adults 16 years or
older to allow for comparison with subsistence and Tribal fishers, which also only estimated exposure
for adults. However, as shown in Table Apx 1-2, the highest fish ingestion rate per kilogram of body
weight for the general population is for a young toddler between 1 and 2 years old. While results are not
shown, the exposure estimates for a young toddler are similar to an adult (i.e., within the same
magnitude). The 50th percentile (central tendency) and 90th percentile ingestion rate (IR) for adults is
5.04 g/day and 22.2 g/day, respectively. The ADRs were calculated using the 90th percentile IR. EPA
typically uses the central tendency for chronic exposure estimates. However, EPA considers both the
central tendency and 90th percentile IRs to be reasonable for the general population. The 90th percentile
IR can also capture individuals within the general population that may have higher chronic exposures
but not as high as the subsistence fisher. As a result, EPA used both fish ingestion rates to estimate an
ADD and LADD. Exposure estimates via fish ingestion were calculated according to the following
equation:
Page 222 of 638
-------
Equation 5-18.
Where:
ADR or ADD =
SWC x BAF xIRx CF1 x CF2 x ED
AT XBW
ADR =
Acute Dose Rate (mg/kg/day)
ADD =
Average Daily Dose (mg/kg/day)
SWC =
Surface water (dissolved) concentration (|ig/L)
BAF =
Bioaccumulation factor (L/kg wet weight)
IR
Fish ingestion rate (g/day)
CF1 =
Conversion factor (0.001 mg/|ig)
CF2 =
Conversion factor for kg/g (0.001 kg/g)
ED
Exposure duration (year)
AT =
Averaging time (year)
BW =
Body weight (80 kg)
The years within an age group (i.e., 62 years for adults) was used for the exposure duration and
averaging time to characterize non-cancer risks. For cancer, the years within an age group was also used
for the exposure duration while the averaging time is 78 years (i.e., lifetime).
A BAF is preferred in estimating exposure because it considers the animal's uptake of a chemical from
both diet and the water column. For TCEP, there are multiple wet weight BAF values reported for whole
fish collected from water bodies that contained TCEP (Table 2-2). The modeled surface water
concentrations were converted to fish tissue concentrations using the upper and lower bound of the
BAFs reported in literature: 2,198 L/kg wet weight for walleye (Sander vitreus) collected from the U.S.
Great Lakes (Guo et al.. 2017b) and 109 L/kg wet weight for mud carp collected from an e-waste
polluted pond in China (Liu et al.. 2019a). While Guo et al. (2017b) is the only U.S. study that measured
TCEP concentrations in fish samples and is presumably more representative of subsistence fishers in the
United States, EPA considered BAF values from non-U. S. studies because of uncertainties with
walleye's BAF and subsistence fishers consume more than just one fish species. As a result, BAF from
non-U.S. studies were considered.
Table 5-31 compares the fish tissue concentration calculated from the scenario-specific modeled surface
water concentrations using the two BAFs with measured fish tissue concentrations obtained from
literature. For comparison, Table 5-31 also includes fish tissue concentrations presented in Table 4-1
that were derived from a BCF. The overall range for scenario-specific fish concentrations based on
modeled concentrations is for wet weight, and monitoring studies reported both wet and lipid weight.
While the lipid content was not available to convert from lipid to wet weight, measured fish tissue
concentrations are still several orders of magnitude lower than that derived from modeled surface water
concentrations and BAF or BCF.
Page 223 of 638
-------
Table 5-31. Fish Tissue Concentrations Calculated from Modeled Surface Water Concentrations
and Monitoring Data
Data
Approach
Data Description
Surface Water
Concentration (jug/L)
Fish Tissue
Concentration
(jig/kg)
BAF (2,198) and the maximum
1-day average dissolved water
concentrations from PSC under
harmonic mean flow conditions
Overall range
3.4E01 to 4.8E03
Overall range
7.6E04 to 1.06E07, ww
Modeled
Surface
Water
Concentration
BAF (109) and the maximum 1-
day average dissolved water
concentrations from PSC under
harmonic mean flow conditions
Overall range
3.4E01 to 4.8E03
Overall range
3.8E03 to 5.3E05, ww
BCF and the maximum 1-day
average dissolved water
concentrations from PSC under
7Q10 flow conditions
Overall range
9.6E01 to 1.0E04
Overall range
3.2E01 to 3.4E03, ww
Fish Tissue
Monitoring
Data (Wild-
Caught)
7 studies with over 200 fish
tissue samples collected from 7
countries, including one U.S.
studv bv Guo et al. (2017b)
Only one non-U.S. study
collected water samples
from the same waterbody
and at the same time as the
fish tissue samples. Surface
water concentrations for
that study ranged from
1.5E-02to2.34E-01
Central tendency range for
U.S. study
6.55 to 3.56E01, lw
Overall range among non-
U.S. studies
ND to 2.96, ww
ND to 1.87E02, lw
Table 5-32 presents the exposure estimates for adult general population fish ingestion doses. These
doses were calculated using the modeled scenario-specific surface water concentrations based on the
50th percentile stream flows and two BAFs.
Page 224 of 638
-------
Table 5-32. Adult General Population Fish Ingestion Doses by Scenario Based on a Production Volume of 2,500 lb/year, High-End
Release Distribution, and Modeled Surface Water
Concentrations Based on 50th Percentile
^low of Harmonic Mean
Scenario Name
SWCfl
(Hg/L)
ADR6
(mg/kg-day)
ADD6 (mg/kg-day)
LADD6 (mg/kg-day)
BAF
2,198
BAF
109
BAF
2,198
BAF
109
BAF
2,198
BAF
109
HE
CT
HE
CT
HE
CT
HE
CT
HE
Import and Repackaging
8.62E02
5.25E-01
2.60E-02
1.19E-01
5.25E-01
5.92E-03
2.60E-02
9.49E-02
4.17E-01
4.71E-03
2.07E-02
Incorporation into Paints
and Coatings - 1-Part
Coatings
3.82E03
2.33
1.15E-01
5.29E-01
2.33
2.62E-02
1.15E-01
4.20E-01
1.85
2.08E-02
9.17E-02
Incorporation into Paints
and Coatings - 2-Part
Reactive Coatings
3.46E03
2.11
1.05E-01
4.80E-01
2.11
2.38E-02
1.05E-01
3.81E-01
1.68
1.89E-02
8.31E-02
Use in Paints and
Coatings at Job Sites
2.03E03
1.24
6.13E-02
2.81E-01
1.24
1.39E-02
6.13E-02
2.23E-01
9.82E-01
1.11E-02
4.87E-02
Formulation of TCEP
Containing Reactive
Resin
4.84E03
2.95
1.46E-01
6.71E-01
2.95
3.33E-02
1.46E-01
5.33E-01
2.34
2.64E-02
1.16E-01
Laboratory Chemicals
3.46E01
2.10E-02
1.04E-03
4.78E-03
2.10E-02
2.37E-04
1.04E-03
3.80E-03
1.67E-02
1.89E-04
8.29E-04
" Surface water concentrations based on 50th percentile flow of harmonic mean flow conditions.
b ADR calculated using the 90th percentile fish ingestion rate (22.2 g/day). ADD and LADD were calculated using both the mean (CT) and 90th percentile (HE) fish
ingestion rates, 5.04 g/day and 22.2 g/day respectively. An ADD based on the 90th percentile ingestion rate is the same as an ADR.
Page 225 of 638
-------
5.1.3.4.3 Subsistence Fish Ingestion Exposure
Subsistence fishers represent a PESS group for TCEP due to their greatly increased exposure via fish
ingestion (142.4 g/day compared to a 90th percentile of 22.2 g/day for the general population) (U.S.
EPA. 2000b). The ingestion rate for subsistence fishers apply to only adults aged 16 to < 70 years. EPA
calculated exposure for subsistence fishers using Equation 5-18 and the same inputs as the non-
subsistence fisher except for the ingestion rate. Furthermore, unlike the general population fish ingestion
rates, there is no central tendency or 90th percentile IR for the subsistence fisher. The same value was
used to estimate both the ADD and ADR.
EPA is unable to determine subsistence fisher exposure estimates specific to younger lifestages based on
lack of reasonably available information. The exposure estimates for an adult subsistence fisher in Table
5-33 were calculated using the array of modeled scenario-specific surface water concentrations and
BAF.
Table 5-33. Adult Subsistence Fisher Doses by Scenario Based on a Production Volume of 2,500
lb/year, High-End Release Distribution, and Modeled Surface Water Concentrations Based on
50th Percentile Flow of Harmonic Mean
Scenario Name
SWC"
(fig/L)
ADD, ADR
(mg/kg-day)
BAF 2,198
ADD, ADR
(mg/kg-day)
BAF 109
LADD
(mg/kg-day)
BAF 2,198
LADD
(mg/kg-day)
BAF 109
Import and repackaging
8.62E02
3.37
1.67E-01
2.68
1.33E-01
Incorporation into paints and coatings -
1-part reactive coatings
3.82E03
1.49E01
7.41E-01
1.19E01
5.89E-01
Incorporation into paints and coatings -
2-part reactive coatings
3.46E03
1.35E01
6.72E-01
1.08E01
5.34E-01
Use in paints and coatings at job sites
2.03E03
7.94
3.94E-01
6.31
3.13E-01
Formulation of TCEP containing reactive
resin
4.84E03
1.90E01
9.40E-01
1.51E01
7.47E-01
Laboratory chemicals
3.46E01
1.35E-01
6.70E-03
1.07E-01
5.33E-03
" Surface water concentrations based on 50th percentile flow of harmonic mean flow conditions.
5.1.3.4.4 Tribal Fish Ingestion Exposure
Tribal populations represent another PESS group. In the United States there are a total of 574 federally
recognized American Indian Tribes and Alaska Native Villages and 63 state recognized Tribes. Many
Tribal cultures are essentially synonymous with and inseparable from their lands and resources (Harper
et al.. 2007). The relationship that Tribal people have towards the land is not one-dimensional and
includes but is not limited to hunting and fishing, food gathering, livestock, commerce and economy, art,
education, health care, and social systems (Harris and Harper. 2011). This relationship forms the basis of
Tamanwit (natural law). Such an intricate connection to the land and the distinctive lifeways and
cultures between individual Tribes create many unique exposure scenarios that can expose Tribal
members to higher doses of contaminants in the environment. However, EPA quantitatively evaluated
only the Tribal fish ingestion pathway for TCEP because of data limitations and recognizes that this
overlooks many other unique exposure scenarios.
U.S. EPA (2011a) (Chapter 10, Table 10-6) summarizes relevant studies on Tribal-specific fish IRs that
covered 11 Tribes and 94 Alaskan communities. EPA used the highest mean IR per kilogram of body
Page 226 of 638
-------
weight reported in a 1997 survey of adult members (16 years and older) of the Suquamish Tribe in
Washington. Adults reported a mean IR of 2.7 g/kg-day, or 216 g/day assuming an adult body weight of
80 kg. In comparison, the IRs for the adult subsistence fisher and general population are 142.2 and 22.2
g/day, respectively. A total of 92 adults responded to the survey funded by ATSDR through a grant to
the Washington State Department of Health, of which 44 percent reported consuming less fish/seafood
today compared to 20 years ago. One reason for the decline is restricted harvesting caused by increased
pollution and habitat degradation (Duncan. 2000).
Because current fish consumption rates are suppressed by contamination, degradation, or loss of access,
EPA reviewed existing literature for IRs that reflect heritage rates. Heritage rates refer to those that
existed prior to non-indigenous settlement on Tribal fisheries resources, as well as changes in culture
and lifeways (U.S. EPA 2016b). Heritage IRs were identified for four Tribes, all located in the Pacific
Northwest region, among available literature. The highest heritage IR was reported for the Kootenai
Tribe in Idaho at 1,646 g/day (RidolfL 2016) (that study was funded through an EPA contract). The
authors conducted a comprehensive review and evaluation of ethnographic literature, historical
accounts, harvest records, archaeological and ecological information, as well as other studies of heritage
consumption. The heritage IR is estimated for Kootenai members living in the vicinity of Kootenay
Lake in British Columbia, Canada; the Kootenai Tribe once occupied territories in parts of Montana,
Idaho, and British Columbia. It is based on a 2,500 calorie per day diet, assuming 75 percent of the total
caloric intake comes from fish and using the average caloric value for fish. Notably, the authors
acknowledged that assuming 75 percent of caloric intake comes from fish may overestimate fish intake.
EPA calculated exposure via fish consumption for Tribes using Equation 5-18 and the same inputs as the
general population except for the IR. Two IRs were used: 216 g/day for current consumption and 1,646
g/day for heritage consumption. Similar to the subsistence fisher, EPA used the same IR to estimate both
the ADD and ADR. The heritage IR is assumed to be applicable to adults. For current IR, U.S. EPA
(2011a) provides values specific to younger lifestages, but adults still consume higher amounts of fish
per kilogram of body weight. An exception is for the Squaxin Island Tribe in Washington that reported
an IR of 2.9 g/kg-day for children under 5 years old. That IR for children is nearly the same as the adult
IR of 2.7 g/kg-day for the Suquamish Tribe. As a result, exposure estimates based on current IRs
focused on adults (Table 5-34).
Table 5-34. Adult Tribal Fish Ingestion Doses by Scenario Based on a Production Volume of 2,500
lb/year, High-End Release Distribution, Modeled Surface Water Concentrations Based on 50th
Percentile Flow, and Two Fish Ingestion Rates
Scenario Name
SWC"
(fig/L)
ADD, ADR
(mg/kg-day)
BAF 2,198
ADD, ADR
(mg/kg-day)
BAF 109
LADD
(mg/kg-day)
BAF 2,198
LADD
(mg/kg-day)
BAF 109
Current mean fish ingestion rate reported by the Suquamish Tribe (216 g/day)
Import and repackaging
8.62E02
5.12
2.54E-01
4.07
2.02E-01
Incorporation into paints and coatings -
1-part reactive coatings
3.82E03
2.27E01
1.12
1.80E01
8.93E-01
Incorporation into paints and coatings -
2-part reactive coatings
3.46E03
2.06E01
1.02
1.63E01
8.10E-01
Use in paints and coatings at job sites
2.03E03
1.20E01
5.97E-01
9.57
4.75E-01
Formulation of TCEP containing reactive
resin
4.84E03
2.88E01
1.43
2.29E01
1.13
Page 227 of 638
-------
Scenario Name
SWC"
(fig/L)
ADD, ADR
(mg/kg-day)
BAF 2,198
ADD, ADR
(mg/kg-day)
BAF 109
LADD
(mg/kg-day)
BAF 2,198
LADD
(mg/kg-day)
BAF 109
Laboratory chemicals
3.46E01
2.05E-01
1.02E-02
1.63E-01
8.08E-03
Heritage fish ingestion rate (1,646 g/day)
Import and repackaging
8.62E02
3.90E01
1.93
3.10E01
1.54
Incorporation into paints and coatings -
1-part reactive coatings
3.82E03
1.73E02
8.57
1.37E02
6.81
Incorporation into paints and coatings -
2-part reactive coatings
3.46E03
1.57E02
7.77
1.25E02
6.17
Use in paints and coatings at job sites
2.03E03
9.18E01
4.55
7.30E01
3.62
Formulation of TCEP containing reactive
resin
4.84E03
2.19E02
1.09E01
1.74E02
8.64
Laboratory chemicals
3.46E01
1.56
7.75E-02
1.24
6.16E-02
" Surface water concentrations based 50th percentile flow of harmonic mean flow conditions.
5.1.3.4.5 Incidental Oral Ingestion from Soil
Average Daily Doses (ADD) were calculated for TCEP ingestion using the following formula:
Equation 5-19.
ADD =
C x IR x EF x ED x CF
BW x AT
Where:
ADD
C
IR
EF
CF
BW
AT
Average Daily Dose (mg/kg/d)
Soil Concentration (mg/kg)
Intake Rate of contaminated soil (mg/d)
Exposure Frequency (d)
Conversion Factor (10x 1CT6 kg/mg)
Body Weight (kg)
Averaging Time (non-cancer: ED x EF; cancer: 78 years x EF)
Modeled soil concentrations were calculated from 95th percentile air deposition (see Section 3.3.3.2)
concentrations for 100 m and 1,000 m from a hypothetical facility. These calculations were conducted
for the COM-Paints-USE scenario (LOW PV - 2,500 lb, HE-95th percentile release).
Soil concentrations of 141.2 ng/g were modeled for 2,500 lb production volume, high-end release
estimate for the Incorporation into paints and coatings - 1 part coatings OES using BST (Section
3.3.3.5).
The mean intake rate for children aged 3 to 6 years varied; 41 mg/d was selected for the mean intake rate
and 175.6 was selected for the 95th percentile intake rate (U.S. EPA 2017d). Body weight (18.6 kg) of a
3- to 6-year-old was estimated from the Exposure Factors Handbook (U.S. EPA. 2017d).
Page 228 of 638
-------
Table 5-35. Modeled Soil Oral Doses for Soil Concentrations Estimated from Air Deposition and
Biosolids Application for the 2,500 lb High-End Release Estimates
OES
Distance
(m)
Soil
Concentration
(ng/g)
Average Daily Dose
(Mean Intake)
(mg/kg-day)
Average Daily Dose
(95th Intake)
(mg/kg-day)
Use in paints and coatings
at job sites'1
100
1.14E04
2.51E-05
1.08E-04
1,000
8.65E01
1.9 IE—07
8.17E-07
Incorporation into paints
and coatings - 1-part
coatings
N/A
1.41E02
3.11E-07
1.33E-06
1195th percentile estimates
5.1.3.4.6 Incidental Oral Ingestion from Swimming
The general population may swim in surfaces waters (streams and lakes) that are affected by TCEP
contamination. Modeled surface water concentrations from EFAST 2014 were used to estimate acute
doses and average daily doses due to ingestion exposure while swimming.
The following equations were used to calculate incidental oral (swimming) doses for all COUs, for
adults, youth, and children:
Equation 5-20.
Equation 5-21.
SWC xIRx CF1
ADR = —
BW
SWC x IRx ED x RD x CF 1
ADD ~ BW x AT x CF2
Where:
ADR = Acute Dose Rate (mg/kg/day)
ADD = Average Daily Dose (mg/kg/day)
SWC = Surface water concentration (ppb or pg/L)
IR = Daily ingestion rate (L/day)
RD = Release days (days/yr)
ED = Exposure duration (years)
BW = Body weight (kg)
AT = Averaging time (years)
CF1 = Conversion factor (l.OxlCT3 mg/|ig)
CF2 = Conversion factor (365 days/year)
Page 229 of 638
-------
A summary of inputs utilized for these estimates are present in Appendix II. 2.
Table 5-36. Modeled Incidental Oral (Swimming) Doses for All COUs, for Adults, Youth and Children, for the 2,500 lb High-End
Release Estimate
OESfl
Surface Water Concentration
Adult (> 21 yrs)
Youth (11-15 yrs)
Child (6-10 yrs)
30Q5
Concentration
(^g/L)
Harmonic Mean
Concentration
(^g/L)
ADRpot
(mg/kg-
day)
ADD
(mg/kg-
day)
ADRpot
(mg/kg-
day)
ADD
(mg/kg-
day)
ADRpot
(mg/kg-
day)
ADD
(mg/kg-
day)
Repackaging of import containers
862.129
1,366.528
2.97E-03
1.29E-05
4.61E-03
2.00E-05
2.60E-03
1.13E-05
Incorporation into paints and coatings -
1-part coatings
3,819.444
5,912.114
1.32E-02
5.59E-05
2.04E-02
8.67E-05
1.15E-02
4.89E-05
Incorporation into paints and coatings -
2-part reactive coatings
3,462.800
5,360.066
1.19E-02
5.07E-05
1.85E-02
7.86E-05
1.05E-02
4.43E-05
Use in paints and coatings at job sites
2,029.305
3,216.574
7.00E-03
3.04E-05
1.09E-02
4.72E-05
6.13E-03
2.66E-05
Formulation of TCEP containing
reactive resin
4,844.722
6,245.374
1.67E-02
5.90E-05
2.59E-02
9.16E-05
1.46E-02
5.17E-05
Use of laboratory chemicals
34.555
54.772
1.19E-04
5.18E-07
1.85E-04
8.03E-07
1.04E-04
4.53E-07
11 Table 3-3 provides a crosswalk of industrial and commercial COUs to OES.
Page 230 of 638
-------
5.1.3.4.7 Human Milk Exposure
Infants are a potentially susceptible subpopulation for various reasons including their higher exposure
per body weight, immature metabolic systems, and the potential for chemical toxicants to disrupt
sensitive developmental processes. To determine whether a quantitative analysis of infant exposure to
TCEP via human milk could be informative, EPA considered available exposure and hazard information
for TCEP. Based on its slight lipophilicity and small mass, TCEP has the potential to accumulate in
milk. In fact, available biomonitoring studies demonstrated the presence of TCEP in human milk.
One U.S. study measured a mean wet weight concentration of 0.036 ng/mL among 100 samples
collected from across the country. The detection frequency was 37 percent, and the range was n.d. to 0.8
ng/mL (Ma et al.. 2019). Among non-U.S. studies, the highest concentrations were observed by Kim et
al. (20141 in which TCEP was measured in 89 milk samples collected in three Asian countries
(Philippines, Japan, and Vietnam), ranging from non-detect to 512 ng/g lipid weight, with an average of
0.14 to 42 ng/g. Another study by Sundkvist et al. (2010) collected milk samples from 286 mothers in
Sweden, where concentrations ranged from 2.1 to 8.2 ng/g lipid weight, with a median of 4.9 ng/g. One
study by He et al. (2018a) collected three milk samples in Australia, and concentrations ranged from n.d.
to 0.47 ng/mL wet weight.
The hazard endpoints identified for TCEP (neurotoxicity for acute scenarios; reproductive toxicity for
intermediate/chronic exposure scenarios as well as carcinogenicity) are relevant for the milk pathway
and are protective of effects that may occur in infants as described in Section 5.2. Because TCEP can
transfer to human milk and infants may be particularly susceptible to its health effects, EPA further
evaluated infant exposures through the milk pathway for specific COUs.
EPA considered all maternal groups—occupational, consumer, and general population—when modeling
TCEP concentrations in milk. Maternal doses are presented in Section 5.1 for occupational, Section
5.1.2.3 for consumer, and Section 5.1.3 for general population.
TCEP concentrations in milk were estimated based on the maternal doses using a multi-compartment
physiologically-based pharmacokinetic (PBPK) model identified by EPA as the best available model
(Verner et al.. 2009; Verner et al.. 2008). hereafter referred to as the Verner model. Only chronic (not
acute) maternal doses were considered because the model is designed to estimate only continuous
maternal exposure. For more information on the Verner model, including modeled compartments, data
input requirements, and its system of differential equations, refer to Appendix 1.5.
The Verner model requires all maternal doses to be entered as oral doses. For consumers, CEM 3.2
already provides inhalation estimates as an oral dose. The only adjustment for maternal consumer doses
was to account for body weight differences. CEM 3.2 assumes a body weight of 80 kg, which is less
representative of women of reproductive age because it combines males and females. To derive a dose
representative of women of reproductive age, EPA applied an adjustment factor of 1.21 based on a body
weight of 65.9 kg (80 kg/65.9 kg) (U.S. EPA. 201 la). The body weight of 65.9 kg is for women 16 to 21
years of age. Body weight increases with age for women of childbearing age, thus reducing overall
exposure estimates. As a result, 65.9 kg is the most health protective. Furthermore, only chronic
maternal doses from consumer scenarios were considered because TCEP is primarily found in consumer
articles that are typically used over a long-time frame.
For OESs, high-end inhalation concentrations were converted to oral equivalent doses using the
following equation:
Page 231 of 638
-------
Equation 5-22.
Inhalation Cone x ED x IR
Oral Equivalent Dose =
BW
Where:
Oral Equivalen t Dose
Inhalation Cone
in mg/kg-day
Inhalation concentration (mg/m3)
8-hour TWA (high-end) for workers
Inhalation rate 1.25 m3/hr for workers
Body weight (65.9 kg)
ED
IR
BW
For workers, maternal dermal doses include both chronic (ADD) and subchronic (SCADD). The
SCADD represents repeated exposure for 30 days or more and is used with intermediate exposure
scenarios. Dermal ADD and SCADD from high-end exposure levels for workers without personal
protective equipment (PPE) (i.e., gloves) were used to estimate infant exposure. These values are
presented in Section 5.1 and adjusted by body weight. Inhalation ADD and SCADD were calculated
using Equation 5-23.
Equation 5-23.
For consumers and workers, maternal doses were combined across all exposure routes for each COU:
inhalation (using the oral equivalent dose calculated with Equation 5-22 and Equation 5-23), dermal,
and/or oral routes. For general population, maternal doses were not combined because certain exposure
pathways (i.e., fish ingestion and undiluted drinking water) demonstrated significantly higher doses than
others and will likely be the main driver of risk. EPA focused on these sentinel exposure pathways.
EPA used 30 years as the age of pregnancy throughout the human milk pathway. This parameter is
applicable to chemicals that accumulate over time. TCEP, being only slightly lipophilic and having a
half-life of less than 24 hours, is not expected to accumulate. Initial model simulations that varied the
age of pregnancy confirmed this expectation. A sensitivity analysis also showed that maternal age had a
negligible effect (see Appendix 1.5).
Infant doses are calculated using the modeled TCEP concentrations in milk and milk intake rates
described in the Agency's Exposure Factors Handbook (U.S. EPA. 201 la) for multiple age groups
within the first year of life. The handbook presents a mean and upper (95th percentile) milk intake rate
for each age group, and infant doses were calculated using both ingestion rates. The model estimated an
average dose for each age group and each milk ingestion rate.
Appendix 1.5.4 presents the average infant doses via the human milk pathway for all COUs within each
maternal group, as well as the range of modeled TCEP concentrations in milk.
D xEFxEY
ADD or SCADC =
Where:
D
EF
EY
ATef
ATey
Oral-equivalent inhalation dose from Equation 5-22 (mg/kg-day)
Exposure frequency (days/yr) (22 days/year for SCADD; 250 days/year for ADD)
Working years (1 year for SCADD; 40 years for ADD)
Averaging time for exposure frequency (30 days for SCADD; 365 days for ADD)
Averaging time for exposure years (1 year for SCADD; 40 years for ADD)
Page 232 of 638
-------
5.1.3.4.8 Dietary Exposure (Non-TSCA)
For general population exposure, literature values indicate dietary exposure from all food groups based
on monitoring data (Table 5-37). The exposure dose associated with ingesting food can be derived by
multiplying the concentration of chemical in food by the ingestion rate for that food and dividing by
body weight (U.S. EPA. 1992). Within this overall framework, exposures could be estimated by
grouping all foods and liquids together and using a generic overall exposure factor, disaggregating
discrete food groups, and using food group specific exposure factors, or estimating exposures for unique
food items.
Other EPA programs such as the Office of Pesticides (OPP) estimates exposure from food from using
two distinct pieces of information: (1) the amount of a pesticide residue that is present in and on food
{i.e., residue level), and (2) the types and amounts of foods that people eat (i.e., food consumption).
Residue levels are primarily developed via crop field trials, monitoring programs, use information
including the percent of crop treated, and commercial and consumer practices such as washing, cooking,
and peeling practices. Various sources provide food consumption data, including the USDA's
continuing survey of Food Intake by Individuals (CSFII), the National Health and Nutrition Examination
Survey (NHANES), What We Eat in America (WWEIA). OPP uses the Dietary Exposure Evaluation
Model-Food Commodity Intake Database (DEEM-FCID) model to estimate dietary exposures. (EPA-
HO-OPP-2007-0780-0001; DEEM-FCID).
For this risk evaluation, EPA used available monitoring data to estimate central tendency and high-end
concentrations of TCEP in specific food groups. Figure 5-9 provides the monitoring concentrations of
TCEP in various food groups.
Page 233 of 638
-------
us Wet
NonUS Wet
659041 - Ftla. 1995 - US
5423396 - He et al..
5423396 - He et al..
5423396 - He el al..
5423396 - He el al..
5423396 - He et al..
5423396 - He et al..
5423396 - He et al..
4292130 - Poma el aL.
4292130 - Foma el at.,
4292130 -Poma el al..
4292130 - Poma el al.,
4292130-Poma el al.,
4292130-Poma ei al..
4292130-Poma el al..
4292130 - Poma el al.,
5166285-Poma el al.,
5166285 - Poma et al.,
5166285-Poma ct al.,
5166285-Poma et al.,
5166285-Poma etal
5166285 -Ponia et al.,
5166285 - Poma et al.,
5166285-Poma etal.,
5166285 - Poma et al
2018-AU
2018 -AU
2018-AU
2018-AU
2018-AU
2018-AU
2018-AU
2018- BE
2018 - BE
2018 - BE
2018- BE
2018-BE
2018 - BE
2018 - BE
2018 -BE
2017- SE
. 2017 - SE
2017 -SE
.2017 - SE
,2017-SE
2017 - SE
. 2017 - SE
. 2017 - SE
. 2017-SE
10A-5
fruil
dairy
1 I'ish and shellfish
| grain
I meal
I non-dairy beverages
I other
I vegetables
baby food-infant formula
fats and oils
Lognormal Distribution (CT and 90th percentile)
Normal Distribuiion (CT and 90th percentile)
Non-Detect
w
mA
m
4*
0.001
0.01 0.1 1
Concentration (ng/g)
100
1000
Figure 5-9. Concentrations of TCEP (ng/g) in the Wet Fraction of Dietary from 1982 to 2018
Page 234 of 638
-------
Table 5-37. Concentrations of Foods Found in the Monitoring Literature in ng/g
Food Type
Count of Estimates from
All Studies (n)
Average of Arithmetic Mean
Estimates for All Data
Average of 90th Percentile
Estimates for All Data
Baby food/formula
1(17)
4.0E-01
6.2E-01
Dairy
3(45)
8.7E-02
1.3E-01
Fats and oils
1(10)
2.6
4.0
Fish and shellfish
1(53)
1.4E-01
3.2E-01
Fruit
1(5)
7.5E-02
9.8E-02
Grain
2(19)
2.3E-01
4.9E-01
Meat
2(50)
3.0E-02
4.7E-02
Vegetables
2(24)
1.4E-01
4.8E-01
Other
2(14)
1.9E-01
2.9E-01
Equations
The equation used to calculate the chronic dose for each age group due to dietary exposure of fruits,
grains, vegetables, meat, dairy, fats, and seafood is presented in Equation 5-24 below.
Equation 5-24.
FCxIRx ED
Where:
ADD =
Average Daily Dose used for chronic non-cancer risk calculations due to
ingestion food group (mg/kg-day)
FC
TCEP concentration in food group (mg/g)
IR
Food group ingestion rate by age group (g/kg bw-day)
ED
Exposure Duration
AT =
Averaging Time
An Australian study indicated that more than 75 percent of the estimated daily intake of TCEP came
from dietary ingestion (4.1 out of 4.9 ng/kg bw/day). This study reported that grains (oatmeal, pasta,
bread) contributed 39 percent and non-alcoholic beverages contributed 32 percent of total TCEP intake
(He et al.. 2018b). Poma et al. (2018) measured TCEP in different food groups in Belgium. In total they
found food intake of TCEP to be 207 ng/d and 2.8 ng/kg/day. TCEP was most concentrated in fats (49
ng/d) and grains (49 ng/d), followed by milk (31 ng/d), meat (23 ng/d), and cheese (23 ng/d). Poma et al.
(2018) suggests that the dietary intake was dominated by fats food group because of the inclusion of the
fish oil supplement fat food group, for which a total of 19 g/d was estimated.
5.1.3.5 Exposure Reconstruction Using Human Biomonitoring Data and Reverse
Dosimetry
EPA describes the approach used to estimate doses based on biomonitoring below. TCEP has been
quantified in human samples in hair, nails (Liu et al.. 2016; Liu et al.. 2015). blood serum, plasma (Zhao
et al.. 2017). urine (Figure 5-10), and human milk (see Section 5.1.3.4.7).
Page 235 of 638
-------
NonUS Unadjusted
HI General Population (Background)
5469782 - He ct al., 2018 - AU
5562397 - Basliaensen el al., 2019 - BE
3020426 - Van Den Eecle et al.. 2015 - AU
0
1 1 10
100 1000 10*4
10*5
Concentration (ng/L)
Figure 5-10. Concentrations of TCEP (ng/L) in the Unadjusted Urine from 2015 to 2019
BCEP, a metabolite of TCEP, has been reported in the 2011 to 2014 NHANES data (CDC. 20131 as
well as the peer-reviewed literature (Wang et al.. 2019d; He et al.. 2018a; Dodson et al.. 2014) (Figure
5-11, Figure 5-12).
General Population (Background)
US Creatinine Adjusted ^ Lognormal Distribution (CT and 90«h percentile)
5164613 - Wang et at. 2019-US
US Unadjusted
2533847 - Dodson ct al., 2014 - US
NonUS Unadjusted
5469782-He etal.. 2018-AU
2537005 - Frommc ct al.. 2014 - DE
0.1 I 10 100 1000 10*4 10*5
Concentration (ng/L)
Figure 5-11. Concentrations of BCEP (ng/L) in the Creatinine-Adjusted Urine from 2014 to 2019
Page 236 of 638
-------
Urinary Bis(2-chiloroethyl) phosphate (BCEtP) (creatinine corrected) (2011 - 2014)
CAS Number 3040-56-0
Metabolite of Tns(2-chloroethyi) phosphate (TCEtP)
Geometric mean and selected percentiles of urine concentrations (in pg/g of creatinine) for the U.S. population from the
National Health and Nutrition Examination Survey.
Demographic
Survey
Geometric Mean
50th Percentile
75ft Percentile
90th Percentile
95th Percentile
Sample
Categories
lYears)
(95% CI)
(95% CI)
(95% CI)
(95% CI)
(95% CI)
Size
Total population
11-12
~.491 {.443-.545)
.4® {.441-.553)
.068 (.811-1.11)
2.11 (1.82-2.35)
3.38 (2.96-3.79)
2409
Total population
13-14
D.447 (.396-.505)
.353 (.337-444)
.856(.743-.SB1)
2.03(1.72-2.38)
3.94 (2.74-5.13)
2619
Age 6-11 years
11-12
D.eea (.8D6-1.16)
.865 (.724-1.13)
1.68(1.51-2.14)
4.22 (2.83-5.44)
6.77 (4.22-15.6)
394
Age 6-11 years
13-14
0.555 {.720-1 .02)
.833 {.876-.931)
1.60(1.13-2.12)
4.25 {3.38-5.43)
6.63 (4.97-3.99)
416
Age 12-19 yea's
11-12
0.574 (.433-750)
.527 (.4D4-.690)
123 (.753-1.90)
3.11(1.80-5.15)
5.15 (2.74-9.05)
366
Age 12-19 yea's
13-14
0.516 (.429- 620)
.442 {.sso-.saa)
1.D6 (7S3-1.3S)
2.33(1.70-3.03)
4.46 (2.42-8.77)
423
Age 20+ years
11-12
0.445 (.396- 501)
.45.7 (.396-.524)
.655(743-1.01)
1.37 {1.60-2.09)
2.69 (2.29-3.45)
1629
Age 20+ years
13-14
D.403 {.362-.460)
.349 {.313-.3931
.742 (632-075)
1.37 {1.42-2.31)
3.12 (2.33-4.60)
isoa
Maes
11-12
0.449 (.413-.43Q)
.449 (.400-.506)
.865 (.779-1.02)
2.07(1.77-2.43)
3.28 (2.ES-4.15)
1217
Maes
13-14
0.42 (370-476)
.373 {.322-.406)
.826 (.725-954)
2.01 (1.50-2.43)
3.70 (2.44-5.50)
1336
Fe-naes
11-12
0.534 (.466- 612)
.534 (.464- 621)
1.04 (.879-122)
2.14 {1.82-2.46)
3.41 (2.76-4.48)
1192
Females
13-14
0.476 {.417- 543)
.407 {.350-.467)
.9D9 (742-1.04)
2.08 {1.75-2.41}
3.99 (2.61-5.26)
1313
Mexican Amercans
11-12
0.432 (.347-.669)
.509 I 381- 666)
1.05 (.673-1.61)
2.13(1.46-3.12)
3.12 (1.97-6.71)
266
Mexican Americans
13-14
0.515 {.394-.672)
.477 {.343-.837)
1.01 (.665-1.47)
2.35(1.57-3.03)
3.18(2.43-8.34)
426
NoM-lispanic Blacks
11-12
0.537 {.460-.599)
.517 (.469-.595)
1.10 (.927-129)
2.43(1.87-2.96)
3.79 (3.08-8.23)
666
Non-Hispanic Backs
13-14
0.374 {.321-.435)
.323 {.267-.450)
.732 (.630-887)
1.56(1.18-1.80)
2.41 (1.86-3.17)
578
Non-Hispanic Whites
11-12
0.466 {.407-.535)
.461 (.399-503)
.800(767-1.09)
1.92 (1.61-2.34)
2.98(2.41-3.72)
776
Non-Hispanic Whites
13-14
D.446 (.393-.506)
.379 {.33S-.437)
.857 (.731-1.00)
2.03(1.64-2.44)
4.68 (2.51 -5.58)
1012
All Hispanics
11-12
D.529 {.446-.626)
.523 (.450-613)
1.D9 (.819-1 41)
2.45(1.97-2.84)
3.43(2.52-5.21)
552
All Hispanics
13-14
0.495 {.4D6-.804)
.472 (.371-.535)
.960 (.738-1.36)
2.27 (1.68-275)
3.14 (2.53-3.84)
666
Asans
11-12
0.606 (.512-.716)
.557 (.473-735?
1.29(1.07-1.58)
2.77 {2.11-3.62)
4.76 (2.77-7.£0)
327
Asians
13-14
0.477 {.412-.553)
.442 {.371-.500)
.792 (.606-1.28)
2.33(1.51-3.46)
4.16(2.78-9.34)
291
Figure 5-12. Concentrations of BCEP from NHANES data for the U.S. Population from 2011 to
2014
TCEP has also been detected in personal hand wipes and wristbands (Figure 5-13, Figure 5-14). Xu et
al. (2016^ calculated dermal absorption daily doses at a mean of 0.088 ng/kg/day.
US
5163584 - Phillips ct al. 2018 - US
g General Population (Background)
2343712 - Staplcton et at. 2014 - US
NonUS
4292136 - Larsson el al, 2018 - SE
3357642 - Xu el al.. 2016 - NO
0.1
10 100 1000
Concentration (ng/wipe)
10*4
Figure 5-13. Concentrations of TCEP (ng/wipe) in Surface Wipes from 2014 to 2018
us
5165046 - Gibson ct al.. 2019 - US
| General Population (Background)
55 Non-Deiecl
3361031 - Kile et al.. 2016 - US
*
0.1
'
10 100
Concentration (ng/g)
1000
Figure 5-14. Concentrations of TCEP (ng/wipe) in Silicone Wristbands from 2012 to 2015
TCEP human biomonitoring data were previously extracted from peer-reviewed studies and curated to
produce one set of summary statistics per study. A total of two peer-reviewed studies, resulting in six
Page 237 of 638
-------
datasets with sampling years from 2014 to 2018, reported TCEP data in human hair, human nails, and
human urine for the U.S. general population. Additional data are available for occupational workers and
highly exposed populations (Mayer et al.. 2021; Shen et al.. 2018; Javatilaka et al.. 2017). Researchers
from the CDC measured urine samples for BCEP in 76 members of the general population and 146
firefighters who performed structure firefighting while wearing full protective clothing and respirators.
BCEP was detected in 10 percent of the general population, but the median concentration was too low to
quantify with acceptable repeatability and accuracy. For firefighters, BCEP was detected in 90 percent
of firefighters at a median of 0.86 ng/mL (Javatilaka et al.. 2017). Table 5-38 provides the number of
datasets for the general population and media type in the United States.
Table 5-38. Human TCEP/BCEP U.S. Biomonitoring Datasets by Population,
Type, and Number
Population
Media Type
# of Datasets
General Population
Human Hair
2
General Population
Human Nails
1
General Population (BCEP)
Human Urine
3
Urinary BCEP was selected as a biomarker of exposure for TCEP. Urinary BCEP is a recommended
target for biomonitoring of TCEP (Dodson et al.. 2014). Furthermore, the robust dataset provided by the
NHANES survey that varies results across demographics, age groups, and time and allows for more
confidence in the values calculated by the exposure reconstruction.
Urinary volume and flow can vary between individuals due to differences in hydration status. One
approach to account for this variability is by taking creatinine-adjusted values for urinary concentration.
The NHANES data already provides creatinine adjusted values and more information on this adjustment
can be referenced in their fourth report (CDC. 2013).
Equation 5-25.
C(' y * CYp
DI =
BW * Fue
Where:
DI = Daily intake of the parent compound (mg/kg-day)
CCr = Creatinine adjusted concentration of analyte in urine (mg biomarker/g creatinine)
Cre = Creatinine excretion rate (g creatinine/day)
BW = Body weight (kg)
Fue = Urinary excretion fraction (mg biomarker excreted/mg parent compound intake)
Kinetic data on the metabolism of TCEP is limited. Literature values have suggested a Fue of 0.07 based
on in vitro human liver microsomes (HLM) experiment, and a value of 0.13 based on in vitro human
liver S9 fraction experiment (Van den Eede et al.. 2013).
The creatinine excretion rate was normalized by body weight (in units of mg creatinine per kg
bodyweight per day). Cre can be estimated from the urinary creatinine values reported in biomonitoring
studies {i.e., NHANES) using the equations of Mage et al. (2008). Assessments from Health Canada and
U.S. Consumer Product Safety Commission (CPSC) have used similar approaches to quantifying
creatinine excretion rate (Health Canada. 2020; CHAP. 2014).
Page 238 of 638
-------
To simplify this analysis, a few excretion rates were selected for various age groups (250 mg/day at 3
years old and 1,750 mg/day for a 20-year-old adult male) from the literature (Mage et al.. 2008). The
2013 to 2014 urinary BCEP concentrations were selected as the most recent and representative
concentrations for the U.S. population. Using the geometric mean and the 95th percentile concentrations
from the 2013 to 2014 NHANES data, the daily intakes are estimated in Table 5-39.
Table 5-39. Reconstructed Daily Intakes from Creatinine Adjusted Urinary BCEP Concentrations
from NHANES (2013-2014).
Statistic
Fue
3-year-old Intake (mg/kg-day)fl
20-year-old Intake (mg/kg-day)6
Geomean
0.13
0.119
0.069
95th Percentile
0.13
0.952
0.525
Geomean
0.07
0.221
0.128
95th Percentile
0.07
1.768
0.975
113-year-old has a BW of 13.8 kg, and Cre of 250 mg/d. Used 6-11 year data for NHANES value (0.855 jj.g/g
geomean and 6.83 jj.g/g 95th percentile) because no data for younger lifestages were available.
b 20-year-old has a BW of 80 kg, and Cre of 1,750 mg/d. Used Adult data for NHANES value (0.408 jj.g/g geomean
and 3.12 jj.g/g 95th percentile).
Wang et al. (2019d) similarly calculated exposure doses of 19 volunteers from Albany, NY of the parent
TCEP using creatinine-adjusted urinary concentrations of BCEP. Wang et al. (2019d) found TCEP
doses to range 11.9 (50th percentile) to 38.6 ng/kg-bw/day. Parameters used by Wang et al. (2019d)
included a 0.63 value for Fue based on literature values for bis(l,3-dichloro-2-propyl) phosphate
(BDCIPP), and daily urine excretion values of 20 mL/kg-bw/day and 22.2 mL/kg-bw/day for children.
Nevertheless, Wang et al. (2019d) stratified TCEP exposure doses by gender, ethnicity and age, and
indicated that females (7.82 ng/kg-bw/day) had higher doses than males (4.35 ng/kg-bw/day),
Caucasians (8.52 ng/kg-bw/day) had higher doses than Asians (4.59 ng/kg-bw/day), and individuals
aged 40 and above (9.61 ng/kg-bw/day) had higher doses than lower age groups.
5.1.3.6 Summary of General Population Exposure Assessment
The general population can be exposed to TCEP from inhalation of air; dermal absorption from soils and
surface waters; and oral ingestion of TCEP in drinking water, fish, and soils. Infants can also be exposed
to TCEP via human milk. The sentinel exposure scenario for general population exposures was fish
consumption. Oral ingestion estimates of fish consumption are provided for the general population and
subsistence fishing populations, as well as Tribal populations, with high-end and central tendency BAF
in Table 5-41.
5.1.3.6.1 General Population Exposure Results
Table 5-40 provides a summary of the acute oral exposure estimates for non-diluted and diluted drinking
water. See Section 5.1.3.4.1 for information on dilution factors used to estimate TCEP concentrations at
drinking water. Table 5-41 provides a summary of the chronic oral exposure estimates for non-diluted
and diluted drinking water; drinking water estimates based on landfill leaching to groundwater;
incidental ingestion of ambient waters during swimming general population and subsistence fisherman
fish ingestion estimates; and 50th and 95th percentile soil intakes at 100 and 1,000 m from hypothetical
facilities. Table 5-42 provides a summary of acute and chronic dermal exposures estimates of dermal
exposure to surface water when swimming and exposure estimates of dermal exposure to chronic
concentration of TCEP in soils. Table 5-43 below provide a summary of the relevant acute, chronic, and
lifetime exposures. These summary tables present oral, dermal, and inhalation exposures as a result
environmental releases (air, water, and disposal releases) for the applicable OES.
Page 239 of 638
-------
Table 5-40. General Population Acute Oral Ingestion Estimates for Drinking Water Summary Table
Acute Oral Exposure Estimates (mg/kg day)
OES"
Drinking Water
Drinking Water (Diluted*)
Adult
(>21
Years)
Infant
(Birth to
<1 Year)
Youth
(16-20
Years)
Youth
(11-15
Years)
Child
(6-10
Years)
Toddler
(1-5
Years)
Adult
(>21
Years)
Infant
(Birth to
<1 Year)
Youth
(16-20
Years)
Youth
(11-15
Years)
Child
(6-10
Years)
Toddler
(1-5 Years)
Import
5.5E-02
1.9E-01
4.2E-02
4.2E-02
5.4E-02
6.9E-02
4.5E-05
1.6E-04
3.4E-05
3.4E-05
4.4E-05
5.6E-05
Incorporation
into paints and
coatings - 1-part
coatings
2.4E-01
8.3E-01
1.8E-01
1.8E-01
2.3E-01
3.0E-01
1.5E-04
5.2E-04
1.1E-04
1.1E-04
1.5E-04
1.9E-04
Incorporation
into paints and
coatings - 2-part
reactive
coatings
2.2E-01
7.6E-01
1.7E-01
1.7E-01
2.1E-01
2.7E-01
1.3E-04
4.7E-04
1.0E-04
1.0E-04
1.3E-04
1.7E-04
Use in paints
and coatings at
job sites
1.3E-01
4.5E-01
9.9E-02
1.0E-01
1.3E-01
1.6E-01
1.0E-04
3.7E-04
8.1E-05
8.1E-05
1.0E-04
1.3E-04
Formulation of
TCEP
containing
reactive resin
2.5E-01
8.8E-01
1.9E-01
1.9E-01
2.5E-01
3.1E-01
5.8E-04
2.0E-03
4.5E-04
4.5E-04
5.7E-04
7.3E-04
Use of
laboratory
chemicals
2.2E-03
7.7E-03
1.7E-03
1.7E-03
2.2E-03
2.8E-03
1.8E-06
6.3E-06
1.4E-06
1.4E-06
1.8E-06
2.2E-06
" Table 3-3 provides a crosswalk of industrial and commercial COUs to OES
h A method was derived to incorporate a dilution factor to estimate TCEP concentrations at drinking water locations downstream from surface water release points.
Because no location information was available for facilities releasing TCEP, a dilution factor and distances to drinking water intake was estimated for each relevant SIC
code. Table 5-26 provides the 50th quantile distances and 50th quantile dilution factors for the relevant SIC codes.
Page 240 of 638
-------
Table 5-41. Summary of General Population Chronic Oral Exposures
Oral (mg/kg/day)
OES°
Drinking
Water
(Dilutedd)
Drinking
Water
Drinking
Water (via
Leaching to
Groundwater)
Ambient
Water
(Incidental
ingestion)
Soil Intake
(50th) at
100 m
Soil Intake
(95th) at
100 m
Soil Intake
(50th) at
1,000 m
Soil Intake
(95th) at
1,000 m
Repackaging of import containers
1.67E-08
2.60E-05
N/A
1.29E-05
1.24E-10
5.30E-10
1.58E-12
6.78E-12
Incorporation into paints and coatings -
1-part coatings
6.20E-08
1.15E-04
1.29E-06
5.59E-05
3.89E-09
1.67E-08
3.44E-11
1.47E-10
Incorporation into paints and coatings -
2-part reactive coatings
5.62E-08
1.04E-04
N/A
5.07E-05
5.63E-10
2.41E-09
7.42E-12
3.18E-11
Use in paints and coatings at job sites
3.92E-08
6.11E-05
N/A
3.04E-05
9.15E-06
3.92E-05
4.77E-08
2.04E-07
Formulation of TCEP containing reactive
resin
2.76E-07
1.46E-04
N/A
5.90E-05
6.19E-10
2.65E-09
7.90E-12
3.38E-11
Processing into 2-part resin article
N/A
N/A
1.29E-06
N/A
5.30E-09
2.27E-08
5.41E-11
2.32E-10
Use of laboratory chemicals
6.68E-10
1.04E-06
N/A
5.20E-07
5.94E-09
2.54E-08
6.50E-11
2.78E-10
OES
General Population (GP)
Subsistence Fisher (SF)
Tribes (Current*)
Tribes (Heritage')
BAF 2,198
BAF 109
BAF 2,198
BAF 109
BAF 2,198
BAF 109
BAF 2,198
BAF 109
Import
5.25E-01
2.60E-02
3.37
1.67E-01
1.89E01
9.40E-01
2.95E01
1.46
Incorporation into paints and coatings -
1-part coatings
2.33
1.15E-01
1.49E01
7.41E-01
8.40E01
4.16
1.31E02
6.47
Incorporation into paints and coatings -
2-part reactive coatings
2.11
1.05E-01
1.35E01
6.72E-01
1.18E02
3.77
1.18E02
5.87
Use in paints and coatings at job sites
1.24
6.13E-02
7.94
3.94E-01
6.94E01
2.21
6.94E01
3.44
Formulation of TCEP containing reactive
resin
2.95
1.46E-01
1.90E01
9.40E-01
1.66E02
5.28
1.66E02
8.21
Processing into 2-part resin article
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Use of laboratory chemicals
2.10E-02
1.04E-03
1.35E-01
6.70E-03
1.18
3.77E-02
1.18
5.86E-02
" Table 3-3 provides a crosswalk of industrial and commercial COUs to OES.
h Current fish consumption rate at 216 g/day based on survey of Suquamish Indian Tribe in Washington (see Section 5.1.3.4.4).
c Heritage fish consumption rate at 1,646 g/day based on study of Kootenai Tribe in Idaho (see Section 5.1.3.4.4).
d A method was derived to incorporate a dilution factor to estimate TCEP concentrations at drinking water locations downstream from surface water release points.
Because no location information was available for facilities releasing TCEP, a dilution factor and distances to drinking water intake was estimated for each relevant SIC
code. Table 5-26 provides the 50th quantile distances and 50th quantile dilution factors for the relevant SIC codes.
Page 241 of 638
-------
Table 5-42. Summary Acute and Chronic General Population Dermal Exposures
Dermal (mg/kg/day)
OESfl
Surface Water
(Swimming)
Soil Mud at 100 m
Soil Activity at 100 m
Soil Mud at 1,000 m
Soil Activity at 1,000 m
Repackaging of import containers
6.00E-06
3.93E-07
1.91E-09
5.02E-09
2.44E-11
Incorporation into paints and
coatings - 1-part coatings
2.60E-05
1.23E-05
6.00E-08
1.09E-07
5.30E-10
Incorporation into paints and
coatings - 2-part reactive coatings
2.40E-05
1.78E-06
8.68E-09
2.35E-08
1.14E-10
Use in paints and coatings at job
sites
1.40E-05
2.90E-02
1.41E-04
1.51E-04
7.36E-07
Formulation of TCEP containing
reactive resin
2.80E-05
1.96E-06
9.54E-09
2.50E-08
1.22E-10
Processing into 2-part resin article
N/A
1.68E-05
8.18E-08
1.71E-07
8.34E-10
Use of laboratory chemicals
2.41E-07
1.88E-05
9.16E-08
2.06E-07
1.00E-09
11 Table 3-3 provides a crosswalk of industrial and commercial COUs to OES.
Table 5-43. Summary of General Population Inhalation Exposures
Inhalation (jig/m3)
OESfl
Ambient Air (50th)
Ambient Air (95th)
Repackaging of import containers
4.39E-10
1.12E-09
Incorporation into paints and coatings - 1-part coatings
1.35E-08
3.51E-08
Incorporation into paints and coatings - 2-part reactive coatings
2.29E-09
1.1IE—08
Use in paints and coatings at job sites
3.36E-05
8.21E-05
Formulation of TCEP containing reactive resin
2.52E-09
1.21E-08
Processing into 2-part resin article
1.96E-08
2.72E-08
Use of laboratory chemicals
2.24E-08
3.33E-08
11 Table 3-3 provides a crosswalk of industrial and commercial COUs to OES.
Page 242 of 638
-------
5.1.3.7 Weight of Scientific Evidence Conclusions for General Population Exposure
Sections 5.1.3.2, 5.1.3.3, 5.1.3.4, and 5.1.3.5 summarize the direct and indirect exposure assessment
approaches taken to estimate general population exposures. A judgment on the weight of scientific
evidence supporting the exposure estimate is decided based on the strengths, limitations, and
uncertainties associated with the exposure estimates. The judgment is summarized using confidence
descriptors: robust, moderate, slight, or indeterminate confidence descriptors.
EPA used general considerations (i.e., relevance, data quality, representativeness, consistency,
variability, uncertainties) as well as chemical-specific considerations for its weight of scientific evidence
conclusions.
EPA modeled three routes of exposure: (1) inhalation from ambient air; (2) oral ingestion from drinking
water, fish ingestion, soil intake, and human milk intake; and (3) dermal exposures from surface water
and soil. Within each of these modeled pathways, EPA considered multiple variations in its analyses
(i.e., multiple distances for inhalation exposures, diluted vs. non-diluted conditions for drinking water
exposures, high vs. low BAF for fish ingestion) to help characterize the general population exposure
estimates and to explore potential variability. The resulting exposure estimates were a combination of
central tendency and high-end inputs for the various exposure scenarios. Modeled estimates were
compared with monitoring data to evaluate overlap, magnitude, and trends. Table 5-44 indicates the
confidence EPA has in their general population exposure estimates for each scenario.
Table 5-44. Overall Confidence for General Population Exposure Scenarios
Route
General Population Exposure Scenario
Confidence
(+ Slight, ++ Moderate, +++ Robust)
Oral
Drinking Water (diluted)
+++
Oral
Drinking Water
++
Oral
Drinking Water (via Leaching to Groundwater)
++
Oral
Surface Water (incidental ingestion)
++
Oral
Fish Ingestion (SF-HighBAF)
+
Oral
Fish Ingestion (GP-HighBAF)
+
Oral
Fish Ingestion (Tribal-HighBAF, Current or Heritage
Ingestion Rate)
+
Oral
Fish Ingestion (SF-LowBAF)
++
Oral
Fish Ingestion (GP-LowBAF)
++
Oral
Fish Ingestion (Tribal-LowBAF, Current or Heritage
Ingestion Rate)
++
Oral
Children's Soil Intake (50th) at 100 m
+
Oral
Children's Soil Intake (95th) at 100 m
+
Oral
Children's Soil Intake (50th) at 1,000 m
++
Oral
Children's Soil Intake (95th) at 1,000 m
++
Oral
Human Milk Intake
+
Dermal
Surface Water (swimming)
++
Dermal
Children playing in Mud at 100 m
+
Dermal
Children activities with Soil at 100 m
+
Page 243 of 638
-------
Route
General Population Exposure Scenario
Confidence
(+ Slight, ++ Moderate, +++ Robust)
Dermal
Children playing in Mud at 1,000 m
++
Dermal
Children activities with Soil at 1,000 m
++
Inhalation
Inhalation 100 m - MetCT
++
Inhalation
Inhalation 1,000 m - MetCT
+++
Inhalation
Inhalation 100 m - MetHIGH
++
Inhalation
Inhalation 1,000 m - MetHIGH
+++
5.1.3.7.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for
the General Population Exposure Assessment
No site-specific information was reasonably available when estimating release of TCEP to the
environment. Release estimates were provided for hypothetical sites. As such, there is considerable
uncertainty in the production volume estimate (2,500 lb), and the resulting environmental release
estimates. In addition, there is uncertainty in the relevancy of the monitoring data to the modeled
estimates presented in this evaluation. Manufacturers have begun to phase out the use of TCEP as
demonstrated by the declining production volumes and the introduction of new regulations (e.g.,
California TB 117-2013) that have shifted the use away from TCEP and other organophosphate flame
retardants. For each release scenario, due to the lack of reasonably available information on the
distribution of TCEP across industry sectors, it was assumed that the full production volume of 2,500 lb
was released for each COU. This conservative assumption further contributes to the uncertainty when
characterizing the resulting modeled exposure estimates.
Drinking Water Estimates
Exposure estimates for the diluted drinking water estimates ranged from 0.022 to 9.167 ug/L which is
one to two orders of magnitude greater than the estimates found in the monitoring literature in the US:
average of 4.9 ng/L and 90th percentile of 9.5 ng/L. The modeled estimates are more in line with a study
of drinking water systems from 19 drinking water systems across the US, where the median measured
concentrations of TCEP in finished water was 0.12 ug/L (Benotti et al.. 2009). There is uncertainty
surrounding the distance between release sites and drinking water intake locations. Nevertheless, the
assessment conducted analyses for diluted and undiluted drinking water estimates to account for this
uncertainty. Only 5 percent of surface water samples detected TCEP in the Water Quality Portal (see
Section 3.3.2.4).
The systematic review resulted in only a few cases demonstrating migration of TCEP to groundwater
from suspected landfill leachate (Buszka et al.. 2009; Barnes et al.. 2004; Hutchins et al.. 1984).
Furthermore, there are inherent uncertainties associated with estimating exposures from the transport of
chemicals through various media (e.g., landfill disposal to groundwater to drinking water). In addition,
TCEP was detected in only 2 percent of groundwater samples in the WQP (see Section 3.3.3.7).
EPA has robust confidence in the diluted drinking water estimate, whereas EPA has moderate
confidence in the non-diluted drinking water estimates. EPA has slight confidence in the drinking water
estimates as a result of leaching from landfills to groundwater and subsequent migration to drinking
water wells.
Page 244 of 638
-------
Fish Ingestion Estimates
To account for the variability in fish consumption across the United States, fish intake estimates were
considered for both subsistence fishing populations and the general population. In estimating fish
concentrations, diluted surface water concentrations were not considered. It is unclear what level of
dilution may occur between the surface water at the facility outfall and habitats where fish reside. A
considerable source of uncertainty in the fish ingestion estimates was the selection of a BAF. Two BAFs
were considered (109 and 2,198 L/kg wet weight) due to uncertainties with the high-end BAF value and
to account for various fish species. No monitoring data were available indicating the consumption of fish
containing TCEP. EPA did find very limited monitoring data indicating TCEP concentrations in fish
tissue. The reported wet weight fish tissue concentrations in the monitoring data are several magnitudes
lower than the modeled estimates with either the low or high BAF.
Soil and Swimming Ingestion/Dermal Estimates
Two scenarios (children playing in mud and children conducting activities with soil) captured a wider
range of potential exposures to TCEP containing soils. EPA's Exposure Factors Handbook provided
detailed information on the child skin surface areas and event per day of the various scenarios (U.S.
EPA. 2017d). It is unclear how relevant dermal and ingestion estimates from soil exposure are as TCEP
is expected to migrate from surface soils to groundwater. Furthermore, there are inherent uncertainties
associated with estimating exposures from the transport of chemicals through various media (e.g., air to
land and subsequent soil ingestion and dermal absorption).
There are no recorded values of TCEP in soils in the United States. A study in Germany reported highest
concentrations of TCEP in soil, 1 day after snow melt at 23.48 ng/g (Mihailovic and Fries. 2012). The
95th percentile estimated modeled concentrations of soil because of air deposition for the use of paints
and coatings at job sites scenario was 1.14><104 ng/g at 100 m and 8.65X101 ng/g at 1000 m. The foreign
monitoring data is within range of the modeled soil estimates via air deposition. The child playing in
mud scenario assumes that the child will be exposed all over the arms, hands, legs, and feet.
Furthermore, there are uncertainties regarding the relevance of the selected dermal absorption fraction of
35.1 percent as discussed in Section 5.1.2.4.1.
Non-diluted surface water concentrations were used when estimating dermal exposures to adults and
youth swimming in streams and lakes. TCEP concentrations will dilute when released to surface waters,
but it is unclear what level of dilution will occur when the general population swims in waters with
TCEP releases.
Inhalation
Modeled inhalation estimates are provided for a range of general population scenarios: various distances
from the emitting facility (10, 30, 60, 100, 1,000, 2,500, 10,000 m), two meteorology conditions (Sioux
Falls, South Dakota, for central tendency meteorology and Lake Charles, Louisiana, for higher-end
meteorology), central tendency and high-end release estimates for the low production volume (2,500 lb),
and 10th, 50th and 95th percentile exposure concentrations. Because no site-specific information for
TCEP release is available, EPA was unable to identify specific meteorological conditions that were
relevant to the air release.
Furthermore, EPA did not consider indoor to outdoor transfer of TCEP for general population inhalation
exposures. As discussed in Section 3.3.1.2.1, there are uncertainties surrounding the particle vs. gas
phase distribution of TCEP. It is unclear how sensitive this parameter is to the final inhalation and
deposition results. Use of paints and coatings at jobs sites was the OES with the highest modeled
exposure estimates (8.21 x 10-5 ppm or 960 ng/m3) which is four orders of magnitude higher than the
Page 245 of 638
-------
average 90th percentile estimates for US data (3.1x10 1 ng/m3). Where information was unavailable,
EPA relied on AERMOD defaults when estimating inhalation exposures.
Reverse Dosimetry
Exposure estimates via reverse dosimetry provide an estimate of exposure based on biomonitoring
concentrations. Although NHANES provides nationally representative biomonitoring estimates, there is
no way to attribute the sources of TCEP to these biomonitoring estimates. NHANES only provided
urinary BCEP concentrations for the years 2011-2014. It is anticipated that these concentrations have
likely decreased due to the decrease in production volume and phase-out of TCEP and accompanying
shift to other alternatives. In addition, there are modeling uncertainties associated with the reverse
dosimetry calculation of estimating internal TCEP doses from BCEP metabolite concentrations.
Uncertainties include creatinine adjustment and the accuracy of urinary excretion fraction. NHANES
biomonitoring estimates do not differentiate between TSCA and non-TSCA exposures. Hence, the
reverse dosimetry estimates will be an overestimate of the actual exposure levels due to TSCA COUs.
The 95th percentile estimate for TCEP intakes from reverse dosimetry is 1.8 mg/kg/day for children
three years of age and 0.98 mg/kg/d for adults 20 years of age. These reverse dosimetry estimates of
TCEP were within an order of magnitude of the highest general population, low BAF, oral fish intake
estimates (0.33 mg/kg/day for formulation of TCEP containing reactive resins OES). This corroboration
builds confidence in the plausibility of the general population fishing exposure estimates.
Key Variables, Parameters for General Population Assessment
Table 5-45 provides a list of key variables and parameters that influence the general population exposure
assessment. This table presents the sources of uncertainties and variabilities of key parameters for the
different exposure scenarios. For more detail on a comprehensive set of parameters used in the general
population exposure assessment, please see Appendix I.
Table 5-45. Qualitative Assessment of the Uncertainty and Variability Associated with General
Population Assessment
Variable Name
Relevant Section(s) in
Risk Evaluation
Data Source(s)
Confidence
(Robust,
Moderate, Slight)
General population exposure assessment
Environmental release
estimates
3.2
EPA Modeled
+
Environmental monitoring
data
3.3.1.1, 3.3.2.2, 3.3.2.3,
3.3.2.7, 3.3.2.8, 3.3.3.1,
3.3.3.3, 3.3.3.6, 3.4.1.1,
3.4.1.2, 3.4.2.1
Extracted and evaluated data (all)
plus key studies
++
Fish intake rate
5.1.3.4.2
(U.S. EPA. 2014a).
(U.S. EPA. 2011a).
(Ridolfi. 2016)
++
Exposure factors and activity
patterns
Appendix I
Exposure Factors Handbook
(U.S. EPA. 2017d)
+++
Key parameters for modeling environmental concentrations
Water modeling defaults:
river flow, dimensions,
characteristics
3.3.2.5, Appendix I
EFAST/VVWM - PSC defaults
++
Page 246 of 638
-------
Variable Name
Relevant Section(s) in
Risk Evaluation
Data Source(s)
Confidence
(Robust,
Moderate, Slight)
General population exposure assessment
Air modeling defaults:
meteorological data,
indoor/outdoor transfer,
3.3.1.2, Appendix I
IIOAC/AERMOD defaults
++
Landfill leachate
concentrations and landfill
loading rates
3.3.3.8
DRAS defaults. (Masoner et al..
2016; Masoner et al.. 2014b)
+
Drinking water treatment and
wastewater treatment removal
2.2.2, F.2.5.2, F.2.5.3
(Life Sciences Research Ltd.
1990b.c)
(Padhve et al.. 2014; Benotti et
al.. 2009; Snvder et al.. 2006;
Westerhoff et al.. 2005;
Stackelbera et al.. 2004).
++
BAF
2.2, 5.1.3.4.2
(Guo et al.. 2017b) and (Liu et al..
2019a).
+ (high BAF)
++ (low BAF)
Gas phase vs. particulate
phase distribution, particle
size
3.3.1.2.1, Appendix I
(Okeme. 2018). (Wolschke et al..
2016).
++
Human biomonitoring and reverse dosimetry parameters
Biomonitoring data
5.1.3.5
Extracted and evaluated data (all)
plus key studies
++
Fraction of urinary excretion
5.1.3.5
(Van den Eede et al.. 2013).
++
Half-life in the body
Appendix I
https://comptox.epa.eov/dashboar
++
d/chemical/adme -ivive -
subtab/DTXSID5021411
Finally, EPA did not consider all possible exposure pathways, but rather focused on pathways that were
within the scope of its conceptual model and most likely to lead to exposures for the general population.
This may result in a potential underestimation of exposure in some cases. Examples of exposure
pathways that were not considered include incidental ingestion of suspended sediment and surface water
during recreational swimming and ingestion of non-fish seafood such as aquatic invertebrates or marine
mammals. However, EPA expects these exposures to be less than those that were included in the overall
assessment for the general population. As such, their impact will likely be minimal and would be
unlikely to influence the overall magnitude of the results.
5.1.3.7.2 Strengths, Limitations, and Key Sources of Uncertainty for the Human
Milk Pathway
Strengths of the Milk Model and Overall Approach
The Verner model integrates critical physiological parameters that includes pre- and postpartum changes
in maternal physiology, lactation, and infant growth. In addition, EPA implemented the Verner Model in
"R" to readily enable adjustments tailored to risk evaluation needs. For example, risk assessors can tailor
model inputs such as maternal doses to be more representative of women of reproductive age, thus
reducing the potential for underestimating infant doses. The overall approach to analyze infant exposure
Page 247 of 638
-------
through human milk also considers a wide range of data sources. It incorporates (1) available
biomonitoring data (see Section 5.1.3.4.7) on TCEP's potential transfer to human milk and its effects on
infants or development, (2) chemical properties influencing TCEP excretion in human milk, and (3) the
best available quantitative approaches for exposure. The half-life for TCEP was estimated using high-
throughput toxicokinetics, which predicts in vivo behavior based on in vitro measures from human
hepatocytes and plasma using simple toxicokinetics model (Wambaugh et al.. 2019). These
considerations were integrated into EPA's decision to proceed with a quantitative exposure analysis.
Uncertainty Associated with Predicting Accumulation in Milk
Well established criteria exist for predicting passive transport of chemicals across cell membranes,
including size, lipophilicity, water solubility, acid/base properties, and ionization. Nevertheless,
predictions of chemical accumulation via passive transport may be confounded by the pH gradient
between plasma and milk. The pH of human milk (7.08) is lower than plasma (7.42). Chemicals that are
weak acids or bases may accumulate to higher levels in milk than predicted based on passive diffusion
due to the pH gradient. For chemicals, the pH change can modify the molecular structure in a manner
that retards diffusion into the plasma medium that is more basic (Alonso-Amelot 2018; Wang and
Needham. 2007). It is not known if TCEP is subjected to ionization trapping because of the pH gradient.
Furthermore, it is not known whether TCEP is a substrate for active transporters in mammary epithelial
cells. These gaps in could introduce uncertainties in how much TCEP accumulates in milk, and thus an
infant's level of exposure.
Uncertainty in the Multi-compartment PBPK Model Inputs and Outputs
The uncertainties associated with deriving maternal doses for workers, consumer, and general
population scenarios are described in Sections 5.1.1.4.1, 5.1.2.4.1, 5.1.3.7.1, respectively. Furthermore,
the model requires oral maternal doses. However, exposure can occur through oral, dermal, and
inhalation pathways for workers, consumers, and the general population. While an inhalation-to-oral
extrapolation of exposures was performed for TCEP to run the model, differences in absorption potential
and/or surface area between the lungs and gastrointestinal tract can introduce uncertainties into the
modeled TCEP concentrations in milk. Also, enzymes involved in xenobiotic metabolism are variably
expressed across many organs and tissues, including sites of absorption such as the gastrointestinal tract,
lung, and skin (Bonifas and B loin eke. 2015; Lip worth. 1996). However, the liver has the highest
detoxification capacity in mammals (Schenk et al.. 2017). After oral administration, xenobiotic
chemicals absorbed from the gastrointestinal tract first pass through the liver before reaching the
systemic circulation. This "first-pass effect" may result in lower systemic bioavailability for chemicals
absorbed via the oral route compared to dermal and inhalation routes (Mehvar. 2018). Therefore, route-
to-route extrapolations may result in underestimating TCEP concentrations in milk. For TCEP, however,
the effect on TCEP concentrations in milk is expected to be small given its relatively slow clearance rate
(i.e., TCEP can partition to other parts of the body because it is not rapidly metabolized by the liver).
The toxicity values used to estimate risks from TCEP exposure are also all based on oral studies (see
Section 5.2.7), and EPA assumed absorption for the oral route is 100 percent (see Section 5.2.5).
Finally, a TCEP-specific source of uncertainty may derive from calculated rather than measured half-life
values and partition coefficients. See Table Apx 1-18 in Appendix 1.5.1 for more information. The
calculated partition coefficients derive from Kow values, lipid and water fractions of blood and tissue,
and previously reported tissue compositions (Verner et al.. 2008; Price et al.. 2003). The lack of
quantifiable uncertainty in these calculated values precludes a robust analysis of their contribution to
overall model uncertainty. However, a sensitivity analysis was conducted for TCEP to evaluate certain
chemical parameters' effects on model estimates. Overall, the model is sensitive to half-life where an
increase or decrease leads to a near equivalent change in the infant milk dose. Kow, which is used to
Page 248 of 638
-------
calculate partition coefficients, has a modest effect on the predicted infant dose. Infant doses are also
insensitive to alterations in milk lipid fraction. Appendix 1.5.1 describes the results of the sensitivity
analysis in greater details.
Uncertainty and Variability Associated with Infant Exposure Dose: The Verner Model assumes
exclusive milk intake for the infant until the end of lactation for up to 12 months. It does not include a
weaning period where formula and/or solid foods are gradually introduced. Therefore, the model may
overestimate infant intake during periods of transition between human milk and formula or solid food
intake.
Weight of Scientific Evidence for Human Milk Pathway
The weight of scientific evidence judgement integrates various considerations to determine confidence
in the evaluation of infant's exposure to TCEP via human milk. The strengths of the Verner PBPK
Model are that it is peer-reviewed and well-documented ("Verner et al.. 2009; Verner et al.. 2008).
However, the model was not validated for TCEP because data were unavailable. It was validated using
data on persistent organic pollutants, which are more lipophilic and have much longer half-lives than
TCEP (i.e., 6-27 years vs. <24 hours) measured in mothers and infants from a Northern Quebec Inuit
population. Furthermore, it is unclear how uncertainties in model inputs like partition coefficients affect
modeled TCEP concentrations in milk. Despite these uncertainties, the Verner PBPK model reflects best
available data identified by EPA, and as such, EPA relied on it to evaluate the human milk pathway.
Biomonitoring data, albeit limited and discussed in Section 5.1.3.4.7, are available to ground truth
modeled concentrations against measured data. Some of the lowest modeled TCEP concentrations in
milk are below measured concentrations, but it is important to note that biomonitoring data does not
distinguish between exposure routes nor allow for source apportionment (i.e., exposure from TSCA
COUs cannot be isolated). EPA has slight confidence in the maternal doses as inputs to the Verner
model human milk. The slight confidence is a result of applying various conservative assumptions in the
absence of site-specific information, route-to-route extrapolation, and other reasons described in
Sections 5.1.1.4.1, 5.1.2.4.1, and 5.1.3.7.1. The infant MOEs based on the modeled concentrations are
still 1 to 2 orders of magnitudes higher than the mothers, in addition to modeled TCEP concentrations in
human milk being higher than measured data for most COUs. Therefore, EPA has moderate overall
confidence that the exposed mothers are more sensitive than infants exposed to TCEP through the
human milk pathway, and therefore protecting the mother is protective of the infant.
5.1.4 Aggregate Exposure Scenarios
EPA has defined aggregate exposure as "the combined exposures from a chemical substance across
multiple routes and across multiple pathways (40 CFR 702.33)." The fenceline methodology (U.S. EPA
2022b). (Draft Screening Level Approach for Assessing Ambient Air and Water Exposures to Fenceline
Communities Version 1.0) aggregated inhalation estimates and drinking water estimates from co-located
facilities. Due to the lack of reasonably available site-specific data for TCEP, EPA was unable to
employ this approach.
Source attribution is a key challenge when attempting to characterize an aggregate exposure scenario.
When considering pathway specific estimates and aggregate exposures, there is uncertainty associated
with which pathways co-occur in each population group. Further, there is variability within a given
exposure pathway. For the same exposure scenarios, central tendency estimates are more likely to occur
than high-end estimates.
Page 249 of 638
-------
Aggregate Exposure across Routes
EPA presents total acute and chronic exposure estimates in the consumer assessment (see Section
5.1.2.3 and Appendix J. 1.3). Generally, exposure estimates to consumer articles are dominated by a
single route (i.e., mouthing by infants and children). However, there are cases where aggregate
exposures across routes are important to consider when inhalation, dermal and ingestion estimates are
within similar ranges, and estimating risks from one route of exposure may underestimate the risk to a
consumer COU. The includes Figure 5-15 that aggregates the consumer exposure estimates by route
(inhalation, dermal, ingestion) for each COU, lifestage combination (U.S. EPA 2024b).
Aggregate Chronic Average Daily Doses (CADDs)
TCEPCOUs
bO
J?
ob
E
cu
l/l
o
Q
u
'c
o
-C
u
Dermal
Ingestion
Inhalation
¦I—.ill J il J_
...ll
111
ID
ID
X
X
X
cz.
CZ.
cz.
n>
ro
rt>
n
n
n
Cj
CJ
CJ
TJ
"O
~a
ID
ID
rn
• i
-<
n
_i
L_
Q)
zr
!_L
ISJ
_J
r-j
r-j
Q_
~E_ ~E_ ~E_ ~E_
Q1 Q1 ai Ql
^< ^< ^< ¦<
i/i i/i un li
d d
n n.
d d
QJQjaiQlQllKQIQl QJQJQJQi
3333333ZB533
3 _ _
-<
O
,— — —n n lj ~r —i ^
C — QJ
q- cl zr _
M M M "
° ° ° (ft (ft (ft (ft
9 - > < n 3
n> ro
L/l l/l l/l
l/l l/l
n =r
— CJ
O- ZT
n> [D
£ 3=
rj qj
3 3
fD ro
n 5"
3 zr zi
l/l LI l/l
— E. E. E E E E
[J [j [j LFl l/l l/l l/l
o o
-< n
o zr
C —
— " a
KJ
rr M
O n n n n fD fD fD
3 n n n n H
— zr zr
_ qT 3 5" 5" 5"
a- zr Cra, Cm Crq (fa
M w > ^ O T
£ £ £ £ £ £ £
o o o o o o o
o o o o o o o
a a cl a a cl cl
l/l L/l l/l l/l
Qj" ST qT rj
zr zr zr 3-
a a cl a
< Q 5"
n
fD fD fD ro
l/l l/l l/l l/l
> < n 3
a ° j ^
C C - rj
7+ =?¦ =r
COU1 Lifestage
C = DJ
S Q. 3
IVJ
CL =T
" Rj
Figure 5-15. Aggregate CADDs for Each Consumer COU, Lifestage
Figure 5-15 demonstrates that for certain consumer products (outdoor play structures, wood resin and
wooden TV stand), exposure is not dominated by a single route and that it is important to consider
multiple routes of exposure. Section 5.3.4 further discusses the aggregate risk characterization of these
COUs and the relevant lifestages.
Aggregate Exposure across COUs
A worker may be involved in multiple activities that use TCEP that have varying multiple OESs.
Consumers may have multiple articles at home that contain TCEP. For example, a consumer could
hypothetically have insulation with TCEP and have wooden articles containing TCEP in the home. No
evidence was found suggesting that a single consumer is exposed through multiple consumer COUs.
Due to lack of reasonably available data indicating co-exposures of multiple TCEP containing activities
Page 250 of 638
-------
or products in the occupational and indoor environment, EPA did not assess aggregate exposure across
consumer, commercial, or industrial COUs.
Aggregate Exposure across Exposure Scenarios
A child in the general population may be exposed to TCEP via soil ingestion and drinking water. In the
case of the general population exposure estimates, a production volume of 2,500 lb used to estimate
releases for each individual OES. EPA did not aggregate exposure estimates to the general population
because exposure estimates were based on release estimates assuming a production volume of 2,500 lb
per OES, and an aggregation would double count the production volume. Thus, in the example above
the soil ingestion estimates were based on 2,500 lb per OES, and the drinking water estimate was based
on 2,500 lb per OES. Thus, it could be misleading to aggregate these exposure estimates.
Furthermore, a child may be exposed to TCEP via mouthing of consumer articles as well as via drinking
water, fish ingestion, or inhalation of ambient air. The source of consumer exposure is via the consumer
purchase of finished articles containing TCEP, whereas the source of environmental exposure from soil
is due to the environmental release from a nearby hypothetical facility. EPA did not quantitively assess
aggregate exposure across exposure scenarios because no data was available indicating the co-exposure
of TCEP from multiple exposure scenarios.
5.1.5 Sentinel Exposures
EPA defines sentinel exposure as "the exposure from a chemical substance that represents the plausible
upper bound of exposure relative to all other exposures within a broad category of similar or related
exposures (40 CFR 702.33)." In terms of this risk evaluation, EPA considered sentinel exposures by
considering risks to populations who may have upper bound exposures; for example, workers and ONUs
who perform activities with higher exposure potential, or consumers who have higher exposure potential
or certain physical factors like body weight or skin surface area exposed. EPA characterized high-end
exposures in evaluating exposure using both monitoring data and modeling approaches. Where
statistical data are available, EPA typically uses the 95th percentile value of the available dataset to
characterize high-end exposure for a given condition of use. For general population and consumer
exposures, EPA occasionally characterized sentinel exposure through a "high-intensity use" category
based on elevated consumption rates, breathing rates, or user-specific factors.
EPA varied the general population exposure scenarios to help characterize the risk estimates. Risk
estimates were calculated for diluted and non-diluted drinking water conditions, soil intakes for
children's activities with soil and playing in mud scenario, and inhalation estimates at various distances
from a hypothetical facility. Furthermore, fish ingestion intakes were estimated using a high and low
BAF value for both subsistence fisherman and the general population. The sentinel exposure for these
general population exposure scenarios was fish ingestion for subsistence fisherman and fishers who are
members of Tribes.
The sentinel exposure for the consumer assessments by route were inhalation from building and
construction materials (roofing insulation) for consumers, oral ingestion of TCEP from children's
mouthing of foam seating and bedding products (foam toy blocks), and children's dermal absorption of
TCEP from wood resin products (wood flooring).
Page 251 of 638
-------
5.2 Human Health Hazard
Human Health Hazards (Section 5.2):
Key Points
EPA evaluated the reasonably available information for human health hazards, including
consideration of the potential for increased susceptibility across PESS factors and acute,
intermediate, and chronic exposures to TCEP (see also Section 5.3.3 and Appendix D). The key
points of the human health hazard assessment are summarized below:
• EPA concluded that TCEP is likely to cause neurotoxicity, kidney toxicity, and reproductive
toxicity based on consideration of evidence across epidemiology studies, animal toxicity
studies with apical outcomes, and mechanistic information.
• EPA concluded that TCEP is likely to be carcinogenic when considering human
epidemiological studies, a 2-year rodent bioassay, and mechanistic evidence.
• Laboratory animal studies identified possible susceptible sex/lifestages (1) males for
reproductive toxicity (with adolescents as potentially most susceptible) (moderate evidence);
(2) neurotoxicity, with greater sensitivity among females (and potentially during pregnancy)
(robust evidence); and (3) reproductive/developmental targets resulting in decreased fertility
and viability of offspring (slight to moderate evidence).
• Human epidemiological data show slight evidence for possible effects in susceptible
subpopulations including developmental effects on growth and gestational age in children of
exposed mothers as well as decreases in IQ among children in with lower socioeconomic
status. There are possible sex differences for some developmental outcomes.
• The acute non-cancer endpoint for TCEP was derived from tremors in pregnant female rats in
a developmental neurotoxicity study with aNOAEL of 40 mg/kg-day.
o Human equivalent dose (HED) (daily) = 9.46 mg/kg-day
o Human equivalent concentration (HEC) (continuous) = 51.5 mg/m3 (4.41 ppm),
extrapolated from oral data
o Benchmark margin of exposure (MOE) = 30, based on 10x intraspecies uncertainty factor
(UF) and 3x interspecies UFs
• The intermediate/chronic endpoint for TCEP was derived from reproductive organ effects
(decreases in seminiferous tubule numbers in adolescent male mice) in a 3 5-day oral feeding
study with a BMDL of 21 mg/kg-day.
o HED (daily) = 2.73 mg/kg-day
o HEC (continuous) = 14.9 mg/m3 (1.27 ppm), extrapolated from oral data
o Benchmark MOE = 30, based on 10* intraspecies and 3x interspecies UFs
• The cancer endpoint for TCEP is based on the observation of kidney adenomas or carcinomas
in male rats from a 2-year oral gavage study.
5.2.1 Approach and Methodology
EPA used the approach described in Figure 5-16 to evaluate, extract, and integrate evidence for TCEP
human health hazard and conduct dose-response modeling. This approach is based on the 2021 Draft
Systematic Review Protocol (U.S. EPA 2021a). updates to the systematic review processes presented in
the TCEP Systematic Review Protocol (U.S. EPA. 2024p). and the Framework for Raman Health Risk
Assessment to Inform Decision Making (U.S. EPA. 2014b).
Page 252 of 638
-------
Figure 5-16. EPA Approach to Hazard Identification, Data Integration, and Dose-Response
Analysis for TCEP
For the human health hazard assessment, EPA systematically reviewed data sources identified in the
literature search conducted in 2019. The literature search was updated in 2024 to search specifically for
human epidemiological and inhalation animal toxicity studies to address limited data for these
endpoints; EPA found several epidemiological studies but no new inhalation animal toxicity studies.
EPA first screened titles and abstracts and then full texts for relevancy using population, exposure,
comparator, and outcome (PECO) screening criteria. Studies that met the PECO criteria were then
evaluated for data quality using pre-established quality criteria and metrics. Although EPA used data
quality criteria for many studies, EPA has not developed such criteria for toxicokinetics data other than
dermal absoiption studies. EPA also did not formally evaluate mechanistic studies for data qualtiy but
did consider whether selected genotoxicity studies followed existing guidelines. Following data quality
evaluation, EPA extracted the toxicological information from each evaluated study, including studies
with uninformative quality determinations. The results of data quality evaluation and extraction of key
study information for dermal absorption studies as well as human and animal phenotypic toxicity studies
are presented in supplemental files CJ.S. EPA 2024q. s, % z).
EPA considered studies that received low, medium, or high overall quality determinations for hazard
identification, evidence integration, and dose-response analysis; only one part of the dermal absorption
study was low quality. Information from studies of uninformative quality were only discussed on a case-
by-case basis for hazard identification and evidence integration and were not considered for dose-
response analysis. For example, if an uninformative study identified a significantly different outcome
compared with high- or medium-quality studies and the uninformative rating was not expected to
influence the specific results being discussed, EPA considered the uninformative study for the hazard
outcome being considered.
After evaluating individual studies for data quality, EPA summarized hazard information by hazard
outcome and considered the strengths and limitations of individual evidence streams (i.e., human studies
of apical (phenotypic) endpoints if available, animal toxicity studies with phenotypic endpoints, and
supplemental mechanistic information). The Agency integrated data from these evidence streams to
arrive at an overall evidence integration conclusion for each health outcome category (e.g., reproductive
toxicity). When weighing and integrating evidence to estimate the potential that TCEP may cause a
Page 253 of 638
-------
given human health hazard outcome, EPA uses several factors adapted from Sir Bradford Hill (Hill.
1965). These elements include consistency, dose-response relationship, strength of the association,
temporal relationship, biological plausibility, and coherence, among other considerations. Sections 5.2.3,
5.2.3.2.9, and 5.2.4 discuss hazard identification and evidence integration conclusions for non-cancer
hazard outcomes, genotoxicity information, and cancer, respectively. Section 5.2.4 also presents an
MOA analysis for cancer.
EPA conducted dose-response analysis for the health outcome categories that received a judgment of
likely ("evidence indicates that TCEP exposure likely causes [health effect]") during evidence
integration. The Agency also conducted dose-response analysis for health outcomes that resulted in
suggestive evidence and compared the PODs (i.e., human equivalent concentrations [HECs] or human
equivalent doses [HEDs] divided by UFs for non-cancer effects; IURs or cancer slope factors [CSFs] for
cancer effects) for both likely and suggestive evidence integration conclusions (U.S. EPA. 2024k).
However, EPA only considered the health outcomes and associated specific health effects from the
likely evidence integration judgments to use as toxicity values when estimating risks from exposure to
TCEP.
If supported by statistically and/or biologically significant results and if the dose-response data could be
reasonably modeled, EPA conducted benchmark dose (BMD) modeling. The dose-response assessment,
including selection of studies and chosen PODs, is discussed in Section 5.2.5.
Finally, EPA assigns confidence ratings for each human health hazard outcome chosen for acute,
intermediate, and chronic exposure scenarios. These ratings consider the evidence integration
conclusions as well as additional factors such as relevance of the health outcome (and associated health
effect[s]) to the exposure scenario (acute, intermediate, or chronic) and PESS sensitivity. This overall
weight of scientific evidence analysis is presented in Section 5.2.6.
Throughout each of these human health hazard analysis steps, EPA considered results of previous
analyses, including EPA's Provisional Peer-Reviewed Toxicity Values for Tris(2-chloroethyl)phosphate
(U.S. EPA. 2009) and the 2009 European Union Risk Assessment Report (ECB. 2009).
5.2.2 Toxicokinetics Summary
This section describes the absorption, distribution, metabolism, and elimination (ADME) data available
for TCEP. For full details on toxicokinetics see Appendix K.l. The PBPK model used to estimate doses
to infants ingesting human milk is described in Section 5.1.3.4.7, with details presented in Appendix 1.5.
In Vivo ADME Information
EPA did not identify in vivo human studies that evaluated ADME information for TCEP by any route of
exposure. However, in vivo ADME studies in rats and mice found that radiolabeled TCEP is rapidly and
extensively absorbed following oral dosing (Burka et al.. 1991; Herr et al.. 1991). TCEP is primarily
eliminated in the urine, with more than 75 percent of a dose of 175 mg/kg eliminated within 24 hours for
both rats and mice (Burka et al.. 1991). TCEP distributes widely throughout the body. Herr et al. (1991)
found radioactivity in blood, liver, and brain (including cerebellum, brainstem, caudate, hypothalamus,
cortex, hippocampus, and midbrain) in male and female rats. There was no significant difference in the
amount of TCEP present in blood and all brain regions after 24 hours of exposure (Herr et al.. 1991).
TCEP is predominantly metabolized in the liver in both rats and mice. Metabolites reported by Burka et
al. (1991) were bis(2-chloroethyl) hydrogen phosphate (BCHP, also identified as BCEP); the
Page 254 of 638
-------
glucuronide of bis(2-chloroethyl) 2-hydroxyethyl phosphate (BCGP); and bis(2-chloroethyl)
carboxymethyl phosphate (BCCP).
Minegishi et al. (1988) is an ADME study that tested TCEP (which is identified as TMCEP in the study)
in 5-week-old male Wistar rats. 14C-labeled TCEP concentrations were measured in urine, feces, expired
air, and body after exposure to a single oral dose. Almost 100 percent of the 50 |imol/kg radioactive
dose was recovered -93.48 percent in urine, 5.64 percent in feces, 1.66 percent in expired air, and 0.76
percent in carcass (body). Biliary excretion for TCEP was nearly 25 percent in 48 hours, and the highest
radioactivity was found in the liver of rats treated with TCEP at 168 hours after administration. The
longest half4ife (second phase) was observed in the adipose tissue of rats treated with TCEP (87 hours)
(Minegishi et al.. 1988).
In Vitro Dermal Absorption
Although no dermal in vivo toxicokinetic studies are available, EPA identified Abdallah et al. (2016).
which measured dermal absorption using excised human skin in multiple in vitro experiments conducted
according to OECD TG 428, Skin Absorption: In Vitro Method. The experiments used exposures of
either 24 or 6 hours; acetone or 20 percent Tween 80 (polyoxyethylenesorbitan monooleate) in water as
the vehicle; 500 or 1,000 ng/cm2 application to skin; and finite (depletable) or infinite dose. EPA gave
each of the finite dose experiments overall quality determinations of medium. For the experiment that
claimed to investigate an infinite dose, EPA assigned a low overall quality determination scenario,
because conditions for infinite dosing (use of neat or large body of material) were not met and the results
did not reflect steady-state flux throughout the experiment (e.g., applied dose was depletable).
EPA used the 500 ng/cm2 24-hour finite dose application in acetone (0.005 percent solution) to estimate
absorption for workers because this was the only experiment for which the authors reported absorption
at multiple time points. Because EPA assumes workers wash their hands after an 8-hour shift, EPA used
the value of 16.5 percent, which is the amount of TCEP absorbed at 8 hours. In accordance with OECD
Guidance Document 156 (OECD. 2022). EPA also added the quantity of material remaining in the skin
(6.8%) at the end of the experiment as potentially absorbable.4 Therefore, EPA assumes workers absorb
23.3 percent TCEP through skin and used this value to calculate risks for workers (see Section 5.1.1.3).
For consumer exposures and exposure to soil scenarios that assume hand washing does not occur for 24
hours, EPA used the value at 24 hours (28.3%) plus the amount remaining in skin (6.8%) from the same
experiment used for workers (500 ng/cm2 24-hour finite dose application in acetone); total absorption
was 35.1 percent absorption and was used to calculate risks (see Sections 5.1.2.2.3 and 5.1.3.3.2).
The estimates identified above apply to finite exposure scenarios for which the TCEP dose is depleted
over time. For exposure scenarios such as swimming in which a maximum absorption rate is expected to
be maintained (i.e., the dose is not depletable during the exposure duration), EPA used the dermal
permeability coefficient (Kp) of 2.2/ 10 2 cm/h derived by Abdallah et al. (2016) from the experiment
that used the 24-hour 1,000 ng/cm2 TCEP skin application to calculate risks (see Section 5.1.3.3.1).
U.S. EPA (2024s) presents quality determinations for individual experiments conducted by Abdallah et
al. (2016). with EPA comments for each of the data quality metrics. Data extraction tables with details
on methods and results of the experiments are also presented in U.S. EPA (2024s).
4 EPA used 6.8 percent (the total amount remaining in skin after washing) because the authors did not conduct tape stripping.
Page 255 of 638
-------
5.2.3 Non-cancer Hazard Identification and Evidence Integration
The sections below describe adverse outcome and mechanistic data available as well as evidence
integration conclusions for each human health hazard outcome (e.g., reproductive toxicity) that has been
examined and/or observed in TCEP human epidemiological and animal toxicity studies. EPA identified
multiple epidemiological studies relevant to non-cancer endpoints, and laboratory animal toxicity studies
with apical endpoints (primarily oral studies), and some mechanistic studies depending on the hazard
outcome.
Section 5.2.3.1 describes the key adverse outcomes with the most robust findings for TCEP that EPA
considered for POD development (i.e., those with likely evidence integration conclusions). Section
5.2.3.2 presents hazard identification and evidence integration for adverse outcome with weaker
evidence.
Appendix L provides more information on the evidence integration conclusions for the TCEP hazard
outcomes. The 2021 Draft Systematic Review Protocol (U.S. EPA 2021a) describes the general process
of evidence evaluation and integration, with relevant updates to the process presented in the TCEP
Systematic Review Protocol (U.S. EPA 2024p).
5.2.3.1 Key Human Health Hazard Outcomes
The sections below focus on hazard identification and evidence integration of neurotoxicity,
reproductive toxicity, and kidney toxicity, which are the most sensitive key human health hazard
outcomes associated with TCEP. These hazard outcome categories received likely evidence integration
conclusions, and sensitive health effects were identified for these hazard outcomes.
In the risk evaluation, neurotoxicity forms the basis of the POD used for acute exposure scenarios and
reproductive toxicity is the basis of the POD used for intermediate and chronic exposure scenarios.
5.2.3.1.1 Neurotoxicity
Humans
EPA identified several epidemiological studies that measured BCEP in urine (Hernandez-Castro et al..
2023a; Percy et al.. 2022; Percy et al.. 2021). and a study that measured TCEP in dust (Foster et al..
2024). In a longitudinal pregnancy and birth cohort, Percy et al. (2021) examined whether prenatal
exposure of suspected neurotoxicants was associated with any changes in child cognition measures.
Pregnant women at 16- and 26-weeks of gestation and at delivery provided urine samples to measure
concentrations of BCEP; the authors then assessed children's cognition at 8 years. Maternal urinary
BCEP was associated with a modest increase in child full scale IQ (FSIQ) with borderline significance
(P = 0.81 per a ln-unit BCEP increase; 95% CI = 0.00, 1.61). Results were adjusted for maternal race,
income, body mass index, and maternal IQ. Child sex did not significantly modify the association
between maternal urinary metabolite concentration and child intelligence in most of the models (Percy et
al.. 2021). Percy et al. (2021) found no adverse association between BCEP concentration and child
cognitive changes at 8 years of age. Quantitative dose-response requires a significant association;
furthermore, an increase in IQ is not considered to be adverse. EPA assigned an overall quality
determination of high to this study.
Another longitudinal cohort study assessed urinary BCEP concentrations at ages 1-5 years and
association with cognitive abilities at 8 years (Percy et al.. 2022). The BCEP concentration association
with child FSIQ was small and positive (P = 0.48; 95% CI = -0.27, 1.24) but not statistically significant.
In addition, the four IQ Index Scores (perceptual reasoning, verbal comprehension, working memory,
and processing speed, (P = 0.37; 95% CI = -0.36, 1.11, p = 0.15; 95% CI = -0.66, 0.95, p = 0.65; 95%
Page 256 of 638
-------
CI = -0.19, 1.49, P = 0.47; 95% CI = -0.37, 1.30, respectively) had 95 percent confidence intervals
which all included the null. Effect modification by measures of socioeconomic status (SES) like
maternal education, race/ethnicity, household income, and neighborhood deprivation were evaluated.
Children with lower SES experienced decreases in FSIQ for every natural log-unit increase in BCEP
concentrations across multiple measures of SES, such as the effect modification of maternal education
(P = -0.93; 95% CI = -1.77, -0.08, P = 0.01) (Percy et al.. 2022). Although the relationships reported by
Percy et al. (2022) mostly were not statistically significant, the authors observed adverse effects of OPEs
on cognitive abilities for children using more than one measure of low SES. Given the inconsistencies in
the outcomes of this study, it does not support a quantitative dose response analysis. EPA assigned a
high overall quality determination to this study.
Foster et al. (2024) examined whether exposure to organophosphate esters (OPEs) that also included
TCEP in home dust was associated with higher depression and stress levels across prenatal and
postpartum time periods. A large nested prospective cohort of 718 mothers in the CHILD cohort study
measured maternal depression and stress at 18-36 weeks gestation and 6 months and 1 year postpartum.
For TCEP, the authors observed some increases in maternal perceived stress levels after adjusting for
covariates when using two different analyses, but the 95th confidence limits all included the null. EPA
assigned a medium overall quality determination to this study.
Hernandez-Castro et al. (2023a) evaluated associations of prenatal exposures to OPEs that included
TCEP and child neurobehavior. The authors did not find an association between urinary BCEP levels
and neurobehavioral outcomes. EPA assigned a medium overall quality determination to this study.
The epidemiological results mostly did not show significant changes in cognitive functions in children
or maternal depression and stress during gestation were associated with maternal urine concentrations of
BCEP as well as TCEP in home dust. Only Percy et al. (2022) showed children with low SES have
reduced IQ. These inconsistencies and lack of association in most results limit any potential dose-
response analysis.
Laboratory Animals
A review of high-quality acute, subchronic, and chronic studies in both rats and mice demonstrated
neurotoxic effects in both sexes following TCEP exposure.
Effects in Adults: Dosing from one to a few days in multiple studies resulted in several signs of
neurotoxicity. Female Fisher-344 rats administered 275 mg/kg of TCEP via oral gavage in a 1-day
toxicity study exhibited increased brain lesions, seizures, and behavior effects (Tilson et al.. 1990). NTP
(1991b) reported that B6C3F1 mice administered the two highest doses (350 or 700 mg/kg-day) in a 16-
day study exhibited ataxia and convulsive movements during the first three days of dosing. Moser et al.
(2015) identified very slight to moderate tremors within three days of dosing at 125 mg/kg-day in 13
pregnant rats. Finally, pregnant mice administered 940 mg/kg-day TCEP via oral gavage were languid,
prostrate, and exhibited jerking movements during GDs 7 through 14 (Hazleton Laboratories. 1983).
Longer-term studies also resulted in multiple neurotoxic effects. NTP (1991b) administered 0, 22, 44,
88, 175, or 350 mg/kg-day TCEP to rats for 16 weeks. Females exhibited greater sensitivity than males.
During week 4, the highest two doses were accidentally doubled, and female rats showed ataxia,
excessive salivation, gasping, convulsions, as well as occasional hyperactivity. Rats exhibited necrosis
of hippocampal neurons with increased dose-response (8 of 10 females at 175 mg/kg-day; 10 of 10
females at 175 and 350 mg/kg-day; and 2 of 10 ales at 350 mg/kg-day); females also showed changes in
the thalamus. Mice did not exhibit neurotoxicity up to 700 mg/kg-day after 16 weeks exposure to TCEP
(NTP. 1991b).
Page 257 of 638
-------
Female SD rats were administered 0, 50, 100, or 250 mg/kg-day TCEP via oral gavage for 60 days
(Yang et al.. 2018a) and exhibited occasional periods of hyperactivity and periodic convulsions at the
highest dose, as well as learning impairment in the acquisition of the water maze tasks at particularly at
100 and 250 mg/kg-day. Histopathological changes in the hippocampus were observed at the two
highest doses that included apoptosis and necrosis as well as invading inflammatory cells
and calcified or ossified foci in the brain cortex at the highest dose (Yang et al.. 2018a).
In a 2-year high-quality study in which rats were administered 0, 44, or 88 mg/kg-day TCEP via oral
gavage, more than 40 percent of 88 mg/kg-day females exhibited histopathological changes such as
focal gliosis, hemorrhage, mineralization, pigmentation, and hemosiderin in the brain stem and
cerebellum (NTP. 1991b). Similar effects were not seen in male rats (only a six percent incidence of
hemorrhage in the pons vs. none in controls). Male mice exhibited some increase in mineralization of
the thalamus (56 and 52 percent at 175 and 350 mg/kg-day compared with 34 percent in controls) with
no T3nges in brain histology in F0 adult CD-I mice dosed with 700 mg/kg-day TCEP via gavage for
several weeks during a cross-over mating study.
Developmental Neurotoxicity: Moser et al. (2015) assessed neurobehavioral effects and related
hormonal responses in a non-guideline study after dosing pregnant Long-Evans rats from GD 10 through
PND 22 via oral gavage of 0, 12, 40, and 90 mg/kg-day.5 The authors measured brain
acetylcholinesterase (AChE) activity, T3 and T4 levels, as well as brain and liver weights in offspring at
PND 6 and 22. Serum AChE was measured in pups at PND22 (after inhibiting butyl cholinesterase
activity). Liver weight, serum AChE, T3, and T4 of dams were measured when they were sacrificed at
PND22. No changes were observed for these measures except an increase in liver weight relative to
body weight of less than 10 percent in dams.
Multiple neurobehavioral tests were conducted. Using an elevated zero maze to measure anxiety-like
behavior, no variables attained statistical significance for offspring of exposed dams when evaluated at
PNDs 35 to 36 or PND 70 to 71. However, the data were highly variable, which could have precluded
detection of effects (Moser et al.. 2015).
In the functional observational battery (FOB) of the offspring, hindlimb grip strength (PND 29 to 30)
and habituation (PND 29 to 30 and 78 to 79) did not differ from controls. The only significant FOB
domain in rats treated with TCEP was activity (sex by-dose-by-day) (p < 0.03), with only the vertical
activity counts in PND 29 to 30 males showing a dose effect (p < 0.01); post-hoc analysis showed no
differences (Moser et al.. 2015).
Offspring were then evaluated as adults (PND 83-101) and were tested for multiple outcomes in the
Morris water maze. In the spatial training portion, TCEP did not result in changes in learning the
platform position (latency, path length, path ratio); swim speed; or working memory (match-to-place).
However, during the memory test, TCEP showed statistically significant dose-response effects for time
in the correct quadrant and proximity score (p < 0.05), although rats in the 40 and 90 mg/kg-day groups
had a greater preference for the target compared to controls. Testing with a visual platform revealed no
differences in swim speed or latency. The authors observed a few differences in tests of spatial search
pattern, although these apparently did not influence the direct learning and memory measurements.
During the righting reflex evaluated from PND 2 to 4, offspring of high-dose TCEP-treated rats showed
a statistically significant sex-by-day interaction on PND 4 (p < 0.05), but there was no statistically
significant overall sex-by-day-by dose interaction. TCEP exposure was not associated with changes in
5 The highest dose was decreased from 125 to 90 mg/kg-day after 5 days.
Page 258 of 638
-------
locomotion using a motor activity ontogeny (on PNDs 13, 17, and 21) or tests that included a light
transition component (PNDs 27 to 28 and 76 to 77) (Moser et al.. 2015). Overall, Moser et al. (2015)
notes that the behavioral changes do not suggest biologically relevant adverse outcomes or
developmental toxicity. Other than tremors in dams early in the study, no TCEP-related adverse effects
were observed in this study.
In a medium quality prenatal study, Kawashima et al. (1983) evaluated effects of TCEP exposure on
neurodevelopment in pregnant Wistar rats gavaged with 0, 50, 100, or 200 mg/kg-day from GD 7
through 15. Twenty-three percent of the dams (7 of 30) at 200 mg/kg-day died between GD 10 and 14,
with no obvious signs of intoxication after death except slight standing hair and weakness. These dams
ate less than controls at GD 10 and 13. Dams at 50 and 100 mg/kg-day did not exhibit similar effects,
nor did they show decreased body weight/weight gain.
Offspring 6 to 7 weeks old were examined for spontaneous behavior, coordinated movements, pain
perception, hearing, and learning ability (Kawashima et al.. 1983). At 200 mg/kg-day, male offspring
exhibited a statistically significant decrease in numbers of rearing (9.8 vs. 19.3 in controls; p < 0.01),
which is one of three measures of spontaneous behavior. The other measures (numbers of ambulation
and fecal bowls) were not affected. The 200 mg/kg-day male offspring also took longer during the
learning ability test (water maze performance) in the last of four trials (p < 0.05); the previous three
trials were not statistically significantly different from controls but males at 200 mg/kg-day took
consistently longer than controls; male offspring did not exhibit an increase in the number of errors in
the water maze. Female offspring did not exhibit any statistically significant effects in these tests, and no
effects were observed in the other neurological tests in either sex. The limited effects in male offspring
were only seen at the highest dose, which resulted in excessive maternal toxicity.
Mechanistic Information
In a 1-day toxicity study, ICR male mice were administered via intraperitoneal injection a single dose at
concentrations of 0, 50, 100, and 200 mg/kg for 2 hours to evaluate the pharmacological effects of
TCEP. Combined administration of TCEP with psychoactive drugs; stimulants and depressants were
used to analyze the neurochemical mechanism involved in the increased ambulatory activity. Data
revealed that significantly high ambulatory activity was seen after the beginning of the measurement and
decreased gradually after the administration of 200 mg/kg of TCEP. The authors note that these results
suggest TCEP acts as a g-amino butyric acid (GABA) antagonist and not as a cholinergic agonist, and
that TCEP increases ambulatory activity in ICR mice through a GABAergic mechanism (Umezu et al..
1998). The Umezu et al. (1998) study was not considered for dose-response analysis because it is not a
relevant route of exposure, but it adds support to the potential neurotoxic nature of TCEP.
(Yang et al.. 2018a) also conducted an analysis to identify possible biochemical processes and metabolic
pathways affected after chronic exposure to TCEP but found low levels of GABA in TCEP-treated
groups.
The metabolic pathway corresponding to GABA and other compounds provide a hypothesis to explore
the possible neurotoxicity mechanisms. These findings have not been further elucidated by additional
studies and thus are not conclusive regarding a mechanism for neurotoxicity.
Page 259 of 638
-------
Serum cholinesterase activity in female rats was 75 and 59 percent of controls (p < 0.01) at 175 or 350
mg/kg-day, respectively after 16-weeks repeated exposure.6 Serum cholinesterase activity was not
reduced in male rats or in either sex of mice after 16 weeks (NTP. 1991b). Moser et al. (2015) did not
identify changes in brain or serum AChE of offspring after developmental exposure. Although serum
cholinesterase activity may be associated with brain activity, The science policy of OPP (U.S. EPA
2000d) concluded that the overall weight of scientific evidence for serum cholinesterase activity is the
weakest link for brain cholinesterase.
Evidence Integration Summary
EPA identified four epidemiological studies that found changes in cognitive functions in children
associated with maternal urine concentrations of BCEP and TCEP in home dust. Thus, the human
evidence is slight for neurotoxicity because although TCEP was associated with some changes in
outcomes, although there was inconsistency among the studies.
The evidence in animals is robust based on the magnitude and severity of histological changes in the
hippocampus and other regions of the brain, clinical signs of toxicity, and behavioral changes in female
rats. Results across available animal toxicological studies showed changes at the highest dose or
increases in a dose-response manner. Effects in offspring did not show greater effects than adults.
The mechanistic data qualitatively support the evidence of hazard for TCEP; however, the data are
indeterminate for the specific mechanism of TCEP hazard and are not able to be used for dose response.
EPA considers the mechanistic evidence to be indeterminate.
Overall, EPA concluded that evidence indicates that TCEP likely causes neurotoxicity in humans under
relevant exposure circumstances. This conclusion is based on effects from oral studies in rats and mice
with dose levels between 22 and 700 mg/kg-day. Compared with exposure in adults, neurotoxicity is not
expected to be increased after developmental exposure based on a lack of effects in a prenatal/postnatal
study with doses up to 90 mg/kg-day (Table_Apx L-l).
5.2.3.1.2 Reproductive Toxicity
EPA guidance defines reproductive toxicity as a range of possible hazard outcomes that may occur after
treatment periods of adequate duration to detect such effects on reproductive systems (U.S. EPA 1996).
Although reproductive toxicity is often associated with developmental toxicity and cannot be easily
separated, this section describes male and female reproductive system toxicity (e.g., effects on sperm,
hormones) as well as effects on mating and fertility in a mouse continuous breeding study. Other
offspring effects from the continuous breeding study (e.g., decreases in live pups per litter) are described
in Section 5.2.3.2.9. Neurodevelopmental investigations are described more fully in Section 5.2.3.1.1.
Humans
EPA did not identify epidemiological or human dosing studies that directly evaluated reproductive
effects from TCEP exposure in the literature search conducted in 2019. Developmental effects from
epidemiological studies are described in Section 5.2.3.2.9.
Laboratory Animals
Animal toxicity studies that evaluated reproductive effects after TCEP exposure consist of one
reproductive assessment by continuous breeding (RACB) in mice (NTP. 1991a) and several repeated-
6 After 16 days, serum cholinesterase activities in female rats receiving 175 or 350 mg/kg-day were 79.7 and 81.8 percent of
controls, respectively; however, this study received an overall uninfonnative quality determination due to a viral infection.
Page 260 of 638
-------
dose studies that evaluated reproductive organs and hormones in adult and adolescent mice and in adult
rats (Chen et al.. 2015a; NTP. 1991b; Matthews et al.. 1990).
A high-quality RACB study (NTP. 1991a) dosed F0 male and female CD-I mice with 0, 175, 350, or
700 mg/kg-day TCEP for 1 week prior to cohabitation, 14 weeks cohabitation, and 3 weeks in a holding
period; F0 mice were allowed to produce up to 5 litters per breeding pair. After weaning of final litters,
the F0 male and female 700 mg/kg-day groups were crossbred with controls of the opposite sex to
determine influence of sex on reproductive outcomes. F1 animals in the final litters of the continuous
breeding phase received TCEP at the same doses as their parents for approximately 14 weeks (from
weaning through 74 days of age, during a 1-week cohabitation phase, and during gestation and
lactation). The F1 animals were then evaluated for reproductive outcomes.7 Because F0 breeding pairs
produced no litters at 700 mg/kg-day, F1 dose groups were limited to 0, 175, and 350 mg/kg-day. F0
control and high dose (700 mg/kg-day) and F1 adult mice were examined for changes in reproductive
organs, sperm parameters, and estrous cyclicity.
Reproductive organs8 of F344 rats and B6C3Fi mice were evaluated in NTP 16-day, 16 to 18 week,9 and
2-year studies (NTP. 1991b) that received overall high-quality determinations, except the 16-day rat
study, which was uninformative due to a viral infection. Matthews et al. (1990) reported results of
additional reproductive measurements (e.g., sperm counts) from the 16- to 18-week NTP studies and
received a medium quality determination for the reported endpoints. Chen et al. (2015a). a high-quality
study, evaluated the male reproductive system at 0, 100, and 300 mg/kg-day TCEP for 35 days in an oral
feeding study of 5-week-old adolescent male ICR mice. U.S. EPA (2024q) presents details extracted
from these studies.
Reproductive Outcomes from RACB: The F0 continuous breeding phase of NTP (1991a). resulted in
decreased fertility;10 values of 72 percent fertility in the fifth litter per breeding pair at 350 mg/kg-day
and 67 to 0 percent in the second through fifth litters at 700 mg/kg-day (p < 0.05) contrasted with F0
control fertility of 97 percent. The 700 mg/kg-day dose also resulted in 25 or more cumulative days to
litter 11 vs. controls beginning in the second litter (p < 0.05).
During crossbreeding of F0 mice, the 700 mg/kg-day male x control female group resulted in lower
pregnancy12 and fertility indices (p < 0.05) but not when treated females were bred with untreated
males.1314 F1 breeding (both sexes dosed) resulted in decreased fertility at 350 mg/kg-day (highest dose;
p < 0.05).
7 The exposure duration was not clearly stated in NTP (1991a) for the F1 generation but Heindel et al. (1989) states that the
continuous breeding protocol specifies that dosing of the F1 generation begins just after weaning.
8 Gross necropsy and histopathology: Males - epididymis, preputial gland, prostate, seminal vesicles, testis; Females -
clitoral gland, mammary glands, ovaries, uterus.
9 NTP (1991b) stated that male rats were dosed for 18 weeks but Matthews et al. (1990) identified the studies as 16-week
studies (vs. an 18-week study for male rats), even though they are the same studies described in NTP (1991b).
111 The percent of mated females with copulatory plugs that got pregnant.
11 This appears to be a measure of the number of days from start of cohabitation of the breeding pairs to the day when pups
were born.
12 Number of fertile pairs of the total number of cohabiting pairs.
13 The number of breeding pairs examined ranged from 18 to 20 among dose groups.
14 NTP (1991a) cited an inhalation study (Shepel'skaia and Dvshginevich. 1981) that administered TCEP at 0, 0.5, and 1.5
mg/m3 to male rats continuously for four months and then mated with unexposed females. Similar to the RACB results, dams
had significantly decreased litter size and also exhibited increased pre- and post-implantation loss at 1.5 mg/m3. Shepel'skaia
and Dvshginevich (1981) appears to be an abstract in Russian; EPA could not obtain this study or evaluate its quality.
Page 261 of 638
-------
Decreased fertility appeared earlier in the second generation (i.e., in the single litters produced according
to protocol) than in the first generation in which only in the second or subsequent litters from each of the
breeding FO pairs were affected.
Male Reproductive Toxicity: In males, effects on reproductive organs and hormone levels were
identified but differed by study and dose. In adolescent mice, Chen et al. (2015a) found 22 and 41
percent decreases in seminiferous tubule numbers at 100 and 300 mg/kg-day, respectively (p < 0.05) as
well as decreases in Leydig, Sertoli, and spermatogenic cells. The 300 mg/kg-day group also resulted in
a testis weight decrease of 13.6 percent and testicular testosterone decrease of 18 percent (p < 0.05) as
well as "absolute" disintegration of seminiferous tubules.
The RACB study (NTP. 1991a) identified a 34 percent decrease in epididymal sperm density, more than
3.4-fold increase in abnormal sperm, 45 percent fewer motile sperm, and a 30 percent decrease in testis
weight (p < 0.001) for the only tested dose (700 mg/kg-day) in the F0 adult CD-I mice. The treated F0
mice also exhibited minimal to mild testes hyperplasia (3/10 vs. 0/10 in controls). F1 male mice did not
exhibit effects on sperm or reproductive organs at either 175 or 350 mg/kg-day (NTP. 1991a).
In the 16-week repeated dose study B6C3Fi mice at 700 mg/kg-day exhibited decreases in absolute and
relative testes weights (p <0.01) (NTP. 1991b). Matthews et al. (1990) reported that the 700 mg/kg-day
mice in this study had slightly reduced sperm counts (p = 0.05). Neither effect was observed at 175
mg/kg-day or lower. No changes in testes weights were observed in male rats up to 175 mg/kg-day after
16 weeks (NTP. 1991b). and sperm morphology could not be conducted on the F344 rats in the 16-week
study due to technical difficulties (Matthews et al.. 1990).15 16 There were no changes in gross necropsy
or histopathology in the 16-day or 16-week NTP studies as identified in the text, or in the 2-year NTP
study as identified in incidence tables (NTP. 1991b).
The crossbreeding results described earlier suggest offspring effects are greater from treated males vs.
treated females.
Female Reproductive Organ and Hormone-Related Effects: Adult F0 females administered 700 mg/kg-
day TCEP in the RACB study exhibited decreased postnatal dam weights but no changes in estrous
cyclicity. Lower doses were not examined, but the treated F1 female adults (175 or 350 mg/kg-day) also
exhibited no estrous cycle changes. Two of ten F1 females at 350 mg/kg-day had ovarian cysts, whereas
none of the ten controls exhibited cysts, although the authors did not suggest this to be a TCEP related
effect.17; lower doses were not evaluated. As noted earlier, even though the RACB identified effects
15 NTP (1991a) provided more details of the sperm morphology and vaginal cytology examinations (SMVCE) from the 16-
week NTP study, citing an unpublished report (Gulati and Russell. 1985) and partly described by Matthews et al. (1990): The
doses evaluated for mice were 0, 44, 175, and 700 mg/kg-day. The 700 mg/kg-day B6C3Fi mice exhibited a 28 percent
decrease in epididymal sperm density; more than a doubling of abnormal sperm; a 22 percent decrease in testicular weight;
and decreased epididymis weights. Rats were evaluated at 0, 22, 88, and 175 mg/kg-day and Gulati and Russell (1985) stated
that rats did not exhibit changes in epididymis and cauda epididymis weights or in percent abnormal epididymal sperm.
Sperm density was reported as being increased and motility was decreased in rats at 175 mg/kg-day even though Matthews et
al. (1990) did not report the results due to technical difficulties. Gulati and Russell (1985) was not readily available;
therefore, EPA did not evaluate it for data quality.
16 In (Shepel'skaia and Dvshginevich. 1981). cited by NTP (1991a). male rats exposed continuously to air concentrations of
TCEP for four months exhibited effects on meiosis, post meiotic growth, and maturity of spennatozoids upon
histopathological examination of males. Shepel'skaia and Dvshginevich (1981) appears to be an abstract in Russian; EPA
could not obtain this study or evaluate its quality.
17 In the F0 700 mg/kg-day dose group, two of 13 females also had ovarian cysts (one minimal, one mild) compared with
none among 12 controls. However, one instance of lymphoma associated with the ovary and one instance of oophoritis was
seen in the controls.
Page 262 of 638
-------
from treated female mice bred with untreated males, effects were less pronounced than those resulting
from treated males crossbred with untreated females (NTP, 1991a).
There were no changes in gross necropsy or histopathology in females in the 16-day or 16-week NTP
studies as noted in the text. No statistically or biologically noteworthy non-cancer effects were seen in
the 2-year study. Although adenocarcinomas occurred in three mice at 350 mg/kg-day (p < 0.05 in the
trend test), a fibroadenoma occurred in control mice; the trend for the combined tumor types was not
statistically significant, and the incidence of adenocarcinoma was within the range of historical controls
(NTP. 1991b).
Mechanistic Information
In vitro studies provide some supporting mechanistic evidence of reproductive effects. Chen et al.
(2015b) identified several effects when mouse Leydig (TM3) cells were exposed to TCEP. At 100
|ig/mL TCEP, which did not result in significant cytotoxicity, effects included large decreases in one
gene associated with testosterone synthesis after all timepoints (6, 12, and 24 hours) and a second gene
at 24 hours. After stimulation of testosterone synthesis genes with human chorionic gonadotropin
(hCG), 100 |ig/mL TCEP still significantly decreased mRNA levels compared with controls or hCG.
Also at 100 |ig/mL and 24 hours exposure, testosterone secretion was decreased by about 50 percent
with TCEP alone and by about 39.9 percent (vs. hCG) after stimulation with hCG. TCEP exposure was
also associated with increased transcription of genes for antioxidant proteins.
Exposure to 300 |ag/mL TCEP (mostly after 24 hours) yielded generally greater changes in
transcriptional levels of genes associated with testosterone synthesis (mostly decreased); increased
transcription of genes encoding antioxidant proteins; increased activities of antioxidants; and decreased
secretion of testosterone. This concentration resulted in 31.4 percent lower viability of cells than
controls; thus, effects at this concentration may be at least partly secondary to cytotoxicity (Chen et al..
2015b). Overall, although some effects may have been due to general cytotoxicity, others are specific to
male reproductive toxicity (Chen et al.. 2015b).
TCEP exposure was not associated with estrogenic or anti-estrogenic effects using either a recombinant
yeast reporter gene assay or by inducing alkaline phosphatase in human endometrial cancer Ishikawa
cells (Follmann and Wober. 2006). Reers et al. (2016) also found no TCEP-related changes in
endogenous androgen receptor (AR) mediated gene expression in metastatic prostate cancer cells
(LNCaP) or in estrogen receptor a (ERa) and the aryl hydrocarbon receptor (AhR) target gene activation
using ECC-1 cells (endometrial carcinoma cells). Krivoshiev et al. (2016) reported that 1,000 |iM TCEP
did not exhibit estrogenic activity in a cell proliferation assay using the breast adenocarcinoma cell line
(MCF-7) but did show anti-estrogenic activity when co-treated with 17P-estradiol (E2), yielding a 32
percent relative inhibitory effect. Viability of TCEP to MCF-7 cells was 93 percent of viability in
controls, and results are not expected to be overly influenced by cytotoxicity.
Evidence Integration Summary
There were no human epidemiological studies available for TCEP for reproductive outcomes (although
see Section 5.2.3.2.9 for developmental outcomes), and the human evidence is indeterminate for
reproductive effects.
For the animal studies, which primarily received high or medium overall quality determinations,
biological gradients were seen for fertility index, number of litters per pair, and number of live pups per
litter, which were decreased in a dose-related manner the F0 generation (NTP. 1991a) and for testes
Page 263 of 638
-------
histopathology in mice (Chen et al.. 2015a). which exhibited increased magnitude and severity with
increasing dose.
Consistent findings included decreased numbers of live pups per litter observed at the same dose in FO
and F1 mice in the RACB, with increasing severity in the second generation (NTP. 1991 a), and decreased
testes weights in mice at 300 mg/kg-day and higher (Chen et al.. 2015a; NTP. 1991a. b). Decreases in
testosterone and related effects were observed in vivo and in vitro (Chen et al.. 2015a; Chen et al..
2015b). with related decreases in gene expression in vitro (Chen et al.. 2015b).
Within and among animal studies, coherent changes were seen between related types of effects.
Decreased testosterone in Chen et al. (2015a) and Chen et al. (2015b) support observed effects on testes
and sperm in other studies. Also, in the first generation of the RACB study (NTP. 1991a). male
reproductive effects were observed along with effects on fertility and live pups per litter.
Some effects differed among studies. Histopathological changes in the testes were also not routinely
identified. Chen et al. (2015a) observed changes in seminiferous tubules in adolescent ICR mice that
were not identified in other studies, including the F1 males in the RACB study that were dosed
beginning at weaning (NTP. 1991a). These differences lend uncertainty regarding the association of this
specific effect with TCEP exposure. However, studies differed in use of species or mouse strains and in
use of gavage vs. feeding. Chen et al. (2015a) was also conducted more than 20 years after the other
studies and differences in assessment methods could possibly explain the differences in results.
Effects on sperm were not identified in the F1 animals even though effects on live pups/litter and
fertility were observed in the RACB study (NTP. 1991a). However, in vitro studies suggest other
mechanisms (e.g., oxidative stress, as suggested by Chen et al. (2015b)) might be operating and could
contribute to the observed reproductive effects.
Overall, evidence in humans is indeterminate based on the lack of available studies. Evidence in animals
is moderate based on studies with decreased testes weight, sperm effects, and/or reduced fertility, and
some support from histopathological changes in testes. EPA considers the mechanistic evidence to be
slight based on decreases in testosterone and gene expression but no direct estrogenic or androgenic
agonism or antagonism. Overall, EPA concluded that evidence indicates that TCEP likely causes
reproductive toxicity in humans under relevant exposure circumstances. This conclusion is based on
effects primarily related to fertility in the RACB study and male reproductive toxicity and is based on
oral studies in rats and mice with dose levels between 22 and 700 mg/kg-day (Table_Apx L-2). EPA
guidelines for reproductive toxicity risk assessment (U.S. EPA. 1996) state that findings in animals are
considered relevant to humans in the absence of evidence to the contrary.
5.2.3.1.3 Kidney Toxicity
Human
EPA identified an epidemiological study that measured BCEP in urine. Kang et al. (2019) examined
whether exposure to OPEs, which also included TCEP, was associated with chronic kidney disease. A
large cohort of 1578 adults who were not currently pregnant had their urine data evaluated for two
chronic kidney disease related parameters which included the estimated glomerular filtration rate
(eGFR) and albumin-to-creatine ratio (ACR). After adjusting for sex, age, race/ethnicity, poverty
income ratio, smoking history, physical activity, and BMI, the association between concentration of a
novel measurement of creatinine-adjusted BCEP levels (ng/mL) and eGFR was negative (P = -1.26) and
statistically significant (p < 0.05); the association was not statistically significant when using the
traditional creatinine adjustment (ng/mg). Urinary BCEP levels showed a marginally significant
Page 264 of 638
-------
association with ACR (P = 0.06; p = 0.08) for both the traditional and novel approaches. The study
authors showed that the methods of urinary dilution adjustment can demonstrate significance and
direction of association. The results show a likely association between TCEP exposure and chronic
kidney disease after adjusting with a novel measurement of creatine. However, the strength of the
association is weak. There is also a lack of consistency in the epidemiology data for the relationship of
TCEP exposure and chronic kidney disease. This is also a cross-sectional study making identification of
causality difficult. Therefore, the results do not support a dose-response analysis. EPA assigned a
medium overall quality determination to this study.
Laboratory Animals
A review of the available animal toxicity studies for rats and mice identified the kidney as the target
organ in both sexes following TCEP exposure. In a short-term (28-day) repeated oral toxicity study,
male Fisher-344 rats were given a daily TCEP dose level of 350 mg/kg-day. Results showed signs of
scattered proximal tubular regeneration in the cortex and outer stripe of the outer medulla (Taniai et al..
2012a). Other findings after short-term exposure included increased absolute and relative kidney
weights in male rats at 175 and 350 mg/kg-day after 16-day oral repeated exposures.
Some effects were also observed after longer-term dosing. After 16 weeks of oral dosing, male rats had
increased absolute and relative kidney weights at high-dose only (350 mg/kg-day) and female rats
exhibited increased absolute and relative weights from 44 to 350 mg/kg-day (NTP. 1991b). Both F0
males and female mice exhibited cytomegaly of renal tubule cells decreased kidney weights and after
dosing of 700 mg/kg-day TCEP for several weeks in a continuous breeding study (NTP. 1991a). In the
16-week study, male mice receiving 700 mg/kg-day had significantly reduced absolute kidney weights,
decreased by 19.4 percent compared to the controls. Relative-to-body kidney weights were decreased at
175, 350, and 700 mg/kg-day by 13.3 percent, 16.0 percent, and 14.1 percent compared to controls.
Tubule epithelial cells with enlarged nuclei (cytomegaly and karyomegaly) were observed in the kidneys
of high-dose (700 mg/kg) male and female mice. These lesions were mostly observed in the proximal
convoluted tubules of the inner cortex and outer stripe of the outer medulla.
In the 2-year bioassay, both sexes of rats and mice exhibited histopathological lesions in the kidney,
including renal tubule hyperplasia and in male and female rats and epithelial cytomegaly and
karyomegaly in both male and female mice (NTP. 1991b).
In the 2-year study, karyomegaly was observed in 32 percent and 78 percent of male mice dosed at 175
and 350 mg/kg-day, respectively, compared to 4 percent of control animals. Karyomegaly was also
observed in 10 percent and 88 percent of female mice dosed at 175 and 350 mg/kg/day, respectively.
Hyperplasia of the renal tubule epithelium was observed in 6 percent and 4 percent of male and female
mice, respectively at 350 mg/kg-day compared to 2 percent and 0 percent of control male and female
mice (NTP. 1991b). High-dose male rats (88 mg/kg-day) exhibited 48 percent incidences of hyperplasia
of the renal tubule epithelium vs. 0 percent in controls. High dose female rats also exhibited increased
incidence of focal hyperplasia of the renal tubule epithelium, by a 32 percent vs. 0 percent in controls
(NTP. 1991b). The authors reported no changes blood urea nitrogen or creatinine in rats or mice.
As noted in Section 5.2.4.2, male rats after two years also exhibited dose-related increased incidence of
renal tubule adenomas vs. control rats (48 vs. 2%); one control and one high dose male developed renal
tubule carcinoma. High-dose female rats exhibited an increased incidence of renal tubule adenomas, but
to a lesser extent than male rats (10 vs. 0 in controls). Eight percent of high-dose male mice had either
renal tubule adenomas or adenocarcinomas compared with two percent in controls.
Page 265 of 638
-------
Mechanistic Information
Mechanistic data also supported the conclusion that TCEP targets the kidney. In a 28-day gavage study,
markers for cell proliferation and apoptosis were increased in the kidneys (OSOM and cortex) of rats
(Taniai et al.. 2012b). In vitro exposure of primary rabbit renal proximal tubule cells (PTCs) resulted in
reduced DNA synthesis, altered expression of cell cycle regulatory proteins, cytotoxicity, inhibition of
ion- and non-ion-transport functions, and there was increased expression of pro-apoptotic regulatory
proteins and decreased expression of proteins that inhibit apoptosis were also observed (Ren et al.. 2012;
Ren et al.. 2009. 2008).
Evidence Integration Summary
EPA identified an epidemiological study that found changes in chronic kidney disease parameters,
estimated glomerular filtration rate (eGFR) and albumin-to-creatine ratio (ACR) associated with urinary
BCEP levels in U.S. adult population.
There was only one human epidemiological study available for TCEP with minimal effects between a
TCEP metabolite and chronic kidney disease parameters and therefore, there is slight human evidence.
The evidence in laboratory animals is moderate based on incidences of kidney histopathology findings
that increased with dose in rats and mice of both sexes. Increased incidences of kidney histopathological
lesions were observed in rats and mice of both sexes following chronic exposures. Although less
consistent, changes in kidney weights were also observed in multiple species. EPA considers the
mechanistic evidence to be slight based on markers of cell proliferation and apoptosis in kidneys of rats
after 28-day gavage treatment and supporting in vitro evidence.
Overall, evidence indicates that TCEP exposure likely causes non-cancer kidney effects in humans
under relevant exposure circumstances based on oral studies with doses ranging from 22 to 700 mg/kg-
day in rats and mice (Table Apx L-4).
5.2.3.2 Other Human Health Hazard Outcomes
This section describes hazard identification and evidence integration for additional non-cancer health
outcome categories not considered to be critical to this risk evaluation based on the results of evidence
integration that identified evidence for these outcomes as suggestive or inadequate to assess effects.
These hazard outcomes are as follows: Skin and eye irritation, mortality, hepatic, immune/
hematological, thyroid, endocrine (other effects), lung/respiratory, and body weight.
5.2.3.2.1 Skin and Eye Irritation
Laboratory Animals
In a medium-quality study (Confidential. 1973). rabbits dermally exposed to 0.5 mL (approximately 279
mg/kg18) TCEP for four hours did not show irritation through 48 hours at either the intact or abraded
skin sites. However, 0.4 mL/kg TCEP (equivalent to 556 mg/kg) was administered to shaved dorsal skin
of rabbits and repeated for four days, resulting in corrosivity and Assuring (FDRL. 1972). This study
received an uninformative overall quality determination based on lack of information on statistical
analysis, and it is not clear how long TCEP was in contact with skin each day or when corrosivity and
fissuring first appeared.
18 According to the accompanying protocol, the dose was 0.5 mL TCEP (equivalent to 695 mg) and some sites were abraded.
Assuming 2.5 kg body weight of rabbits (2 to 3 kg was identified in the accompanying protocol), the dose was approximately
279 mg/kg-bw.
Page 266 of 638
-------
TCEP was not irritating to eyes of rabbits when administered at 0.1 mL and observed for 72 hours
(Confidential 1973) in a medium-quality study.
Evidence Integration Summary
The human evidence is indeterminate for skin and eye irritation. The two readily available dermal
irritation studies in animals showed inconsistent results and the single eye irritation study of medium
quality showed that TCEP is not irritating; these studies are indeterminate. Although one study was
uninformative, EPA considered that these results are not affected by the lack of statistical analysis.
Overall, the currently available evidence is inadequate to assess whether TCEP causes irritation in
humans (see Appendix L.2).
5.2.3.2.2 Mortality
Laboratory Animals
EPA identified multiple oral studies and two dermal studies. In short-term oral mouse studies, no female
CD-I mice died at 940 mg/kg-day after dosing from GD 7 to 14 (Hazleton Laboratories. 1983). Seven of
thirty pregnant female Wistar rats died in a prenatal study (Kawashima et al.. 1983). In a 16-day
repeated-dose study, no mice died at doses up to 350 mg/kg-day (NTP. 1991b).19 At higher doses, 13 to
20 percent female mice died at 1,000 mg/kg-day and all mice died at 3,000 mg/kg-day after 8 to 14 days
of exposure (NTP. 1991a; Hazleton Laboratories. 1983).
In longer-term studies, adult mortality was observed at lower doses in rats compared with mice. In 16- to
18-week subchronic studies that received medium-quality determinations for mortality, male and female
rats exhibited decreased survival as low as 175 and 350 mg/kg-day, respectively, but both groups
accidentally received double doses during week four; no mice died at doses up to 700 mg/kg-day after
16 weeks (Matthews et al.. 1990).20 No deaths occurred in rats or mice at lower doses (250 to 300
mg/kg-day) for 35 or 60 days (Yang et al.. 2018a; Chen et al.. 2015a); both studies received overall
high-quality determinations. In a high-quality 2-year study, rats exhibited decreased survival (by 27 to
29%) at 88 mg/kg-day, but mice did not exhibit differences in survival up to 350 mg/kg-day (NTP.
1991b).
In a medium-quality dermal irritation study, four of six rabbits died after a four-hour exposure to
approximately 279 mg/kg TCEP (Confidential. 1973).21 These rabbits exhibited narcosis and paralysis
before death. However, FDRL (1972) did not report any deaths in rabbits dermally exposed to
approximately 556 mg/kg for 4 days. This study received an uninformative overall quality determination
based on lack of information on statistical analysis.
Decreases in numbers of live born animals after parental exposure are described in Section 5.2.3.1.2.
Evidence Integration Summary
Human evidence is indeterminate for mortality because there are no human epidemiological studies that
assessed this endpoint. There is modest evidence in animal studies that shows higher mortality in rats
than mice on oral studies and uncertain potential for mortality via the dermal route given conflicting
19 No rats died in a short-term study at doses up to 700 mg/kg-day (NTP. 1991b) that received an uninformative overall data
quality determination due to a viral infection.
211 NTP (1991b) reported that 9 of 10 male rats survived at 175 mg/kg-day in the 16-week study compared with 4 of 10
reported by Matthews et al. (1990). which is a report of the same study.
21 The 2009 European Union Risk Assessment Report (ECB. 2009) reported results of an acute dermal study not readily
available to EPA in which four rabbits were each exposed dermally to 2,150 mg/kg for 24 hours, using occlusive patches. No
deaths, apparent signs of toxicity, or cholinesterase depression were observed in any of the rabbits 72 hours after treatment.
Page 267 of 638
-------
results. Overall, evidence suggests but is not sufficient to conclude that TCEP exposure causes mortality
in humans under relevant exposure circumstances. This conclusion is based on oral studies in rats and
mice that assessed dose levels between 12 and 700 mg/kg-day and dermal studies in rabbits at
approximately 279 and 556 mg/kg-day (see Appendix L.2).
5.2.3.2.3 Liver
Laboratory Animals
EPA identified multiple high-quality animal studies that reported liver weight, histopathological
changes, and one study measured enzyme changes. Liver weights were statistically increased in multiple
oral gavage rodent studies. In 16- or 18-week studies, rats and mice exhibited absolute increases ranging
from 10 to 84 percent and relative-to-body weight increases ranging from less than 10 to 51 percent,
with the largest increases in female rats at the highest dose of 350 mg/kg-day (NTP. 1991b).22 At the 66-
week sacrifice in the chronic bioassay, male rat absolute and relative liver weights were increased by 20
and 19 percent, respectively, at 88 mg/kg-day (the highest dose) but female rats did not exhibit similar
changes. Liver weight was not reported for mice in the chronic bioassay (NTP. 1991b).23 F0 male mice
(but not females) given 700 mg/kg-day TCEP for 18 weeks in a continuous breeding study via oral
gavage exhibited increases in relative and absolute liver weight of 20 and 15 percent, respectively, with
no accompanying body weight changes (NTP. 1991a). No liver weight changes were seen after 350
mg/kg-day in the F0 or F1 generation in the same study. Only the 16-day mouse study reported a
decrease in (relative) liver weight in males (by 18%), but the change was seen only at 44 mg/kg-day
without a dose-response (NTP. 1991b).24
In the 2-year oral gavage bioassay, male mice had 6 and 16 percent incidence of eosinophilic liver foci
at 175 and 350 mg/kg-day compared with zero incidence in controls. EPA conducted a Fischer's exact
test and identified the incidence at the highest dose to be statistically significant (p < 0.01). The foci are
believed to be precursors to hepatocellular neoplasms (NTP. 1991b). Because these foci were not
accompanied by increased basophilic and clear cell foci, which are considered part of the continuum
with hepatocellular adenomas, NTP (1991b) states that it is uncertain whether eosinophilic foci were
associated with TCEP exposure. Adenomas and carcinomas are discussed in Section 5.2.4.2. At 700
mg/kg-day in the continuous breeding study, F0 male mice exhibited cytomegaly (10/12) and hepatitis
(4/12) vs. 0/10 per effect in controls; no other doses were evaluated in the F0 generation. F1 mice
exhibited minimal or mild changes in liver histology at 350 mg/kg-day (NTP. 1991a).
Liver enzyme activity was measured only at the 66-week sacrifice in the 2-year bioassay (NTP. 1991b).
Female rats at 88 mg/kg-day exhibited significantly decreased mean serum alkaline phosphatase (ALP)
and alanine transferase (ALT) values with no change in aspartate transaminase (AST). No information
was provided on the magnitude of change, and no differences were reported for male rats or mice of
either sex (NTP. 1991b). Although increases in liver enzyme activity are typically associated with liver
injury, decreases are harder to interpret. Decreases in serum ALT could occur after initial increases
resulting from liver injury and has been associated with decreased levels of vitamin B6 (Giannini et al..
2005). ALP is also present in bone and intestines and decreases have been associated with chronic
22 The 350 mg/kg-day female rats also had increased body weight (by 20%) compared with controls (NTP. 1991b).
23 In the 16-day rat study, females exhibited statistically significant increases in absolute and relative liver weights (by 17 and
14 percent, respectively) at 350 mg/kg-day but the study was uninfonnative due to a viral infection.
24 Chen et al. (2015a) found that male mice had decreases of 17.3 and 18.1 percent in absolute liver weight at 100 and 300
mg/kg-day, respectively after 35 days of dosing in an oral feeding study. Body weights were also decreased by 13.5 and 14.8
percent at 100 and 300 mg/kg-day respectively (estimated from graphs using Grablt!™ Copyright Datatrend Software, 1998-
2001. https://download.cnet.com/Grab-It-XP/3000-2Q53 4-41084.html). EPA calculated decreased liver weights relative to
body weights for male mice of 3.5 and 3.6 percent at 100 and 300 mg/kg-day, respectively (Chen et al.. 2015a): therefore, the
changes were within 10 percent and not considered adverse.
Page 268 of 638
-------
myelogenous leukemia, anemias, severe enteritis, and other conditions (Sharma et al.. 2014; Giannini et
al.. 2005V
Due to uncertainty and lack of reasonably available information, EPA has not determined the decreased
enzyme activities to be adverse. Furthermore, except for the liver weight changes identified in the
reproductive and continuous breeding protocol in male mice at 700 mg/kg-day that were accompanied
by histopathological changes, the increased liver weights in other studies are not clearly adverse due to
the lack of histopathological changes and lack of increased enzyme activity.
Mechanistic Information
EPA identified mechanistic studies in liver and liver cells from both in vivo and in vitro studies. Limited
mechanistic data indicate that TCEP may increase oxidative stress (based on increased hepatic
antioxidant enzyme activities and accompanying gene expression) in the livers of male ICR mice after
35 days of dietary TCEP exposure (Chen et al.. 2015a). In vitro studies show that TCEP induced
oxidative stress, altered cellular energetics, and influenced cell signaling related to proliferation, growth,
and cell survival in the liver (Mennillo et al.. 2019; Zhang et al.. 2017b; Zhang et al.. 2017a; Zhang et
al.. 2016c; Zhang et al.. 2016b).
Evidence Integration Summary
There are no epidemiology studies that investigated liver effects, and human evidence is indeterminate.
Male mice exhibited a dose-related increase in eosinophilic foci after two years (as well as an increase in
hepatocellular adenoma) in a high-quality study (NTP. 1991b). Increases in absolute and relative liver
weights in male and female rats occurred at lower doses as duration increased from 16 days to 16 weeks
(NTP. 1991b). Absolute and relative liver weights also generally increased dose-dependently in female
rats and female mice at 16 weeks and in male rats at 66 weeks (NTP. 1991b). Only at a higher dose (700
mg/kg-day) was concordance observed between increased absolute and relative liver weight and
histopathological changes (NTP. 1991a).
However, NTP (1991b) suggests an uncertain association between TCEP exposure and eosinophilic foci.
Also, there were no histopathology findings in rats or female mice, including no hypertrophy associated
with liver weight increases. Liver weight increases were seen in female rats after 16 days and 16 weeks,
but not 66 weeks of exposure. Increased liver weight was not seen in the 35-day study (Chen et al..
2015a). No biologically relevant changes in serum enzymes were seen in the 2-year bioassay and were
not measured in shorter studies. Therefore, EPA determined that the animal evidence for adverse effects
on the liver based on these data are slight for the association between TCEP and adverse liver effects.
Mechanistic information shows biological gradients for the induction of hepatic oxidative stress
occurring earlier than apical endpoints. Also, across the in vitro studies, dose-related changes in
viability, oxidative stress, and impaired mitochondrial functioning were observed. Oxidative stress is a
plausible mechanism for eosinophilic foci (and tumor formation) that is relevant to humans. However,
few potential mechanisms were investigated in available studies and oxidative stress was demonstrated
in vivo at higher doses than those associated with liver lesions in the chronic study. This information
suggests mechanistic evidence for liver effects is slight.
Based on the indeterminate human evidence, slight animal evidence showing increased liver weights in
in the absence of relevant clinical chemistry findings or statistically significant histopathology changes,
EPA concluded that evidence suggests but is not sufficient to conclude that TCEP exposure causes
Page 269 of 638
-------
hepatic toxicity in humans under relevant exposure circumstances. This conclusion is based on studies
of mice and rats that assessed dose levels between 44 and 700 mg/kg-day (see Table_Apx L-5).
5.2.3.2.4 Immune/Hematological
Humans
In a high-quality study, Mendv et al. (2024) studied associations between multiple exposure measures
(TCEP in dust as concentrations and loadings and BCEP in urine of mothers during 16 and 26 weeks of
gestation and at delivery) for 342 mother-infant pairs in Cincinnati, Ohio and the risk of wheeze,
respiratory infections, and hay fever/allergies. Mothers were given a questionnaire asking about
respiratory symptoms in their children every six months up to five years of age. At five years,
spirometry was assessed in the children, including forced expiratory volume in 1 second (FEVi) and
peak expiratory flow (PEF). FEVi was compared to a referent population of children.
TCEP and BCEP concentrations were treated as continuous variables when comparing with lung
function. Another model investigated changes through time by dichotomizing the concentrations as less
than or greater than the geometric mean.
Respiratory Infections: Urinary BCEP levels at 16 weeks (but not 26 weeks or at delivery) were
associated with respiratory infections (relative risk (RR): 1.43; 95% CI: 1.08-1.90) in a model adjusted
for child's sex, child's race/ethnicity, birth weight, gestational term, family income, and child receiving
breast milk. When dichotomizing concentrations (greater or less than the geometric mean), high BCEP
concentrations at 16 and 26 weeks were associated with respiratory infections (RR: 1.78; 95% CI: 1.04-
3.07). TCEP in dust was not associated with increases in respiratory infections (Mendv et al.. 2024).
Wheeze: When dichotomizing the concentrations, high BCEP concentrations in urine at gestational
weeks 16 and 26 were associated with higher risk of wheeze (RR: 1.63; 95% CI: 1.12-2.36). Similarly,
high BCEP at week 16 of gestation and at delivery was associated with higher risk of wheeze (RR: 1.88;
95% CI: 1.18-2.99) (Mendv et al.. 2024).
Hay Fever Allergies: High BCEP at week 16 of gestation and at delivery was associated with higher risk
hay fever/allergies (RR: 1.65; 95% CI: 1.08-2.51). TCEP dust concentrations at 20 weeks gestation
were associated with a higher risk of hay fever/allergies (RR: 1.11; 95% CI: 1.01-1.21), with a similar
relationship between dust loadings and hay fever/allergies; both dust models were adjusted for child's
sex, child's race/ethnicity, birth weight, gestational term, family income, and child receiving breast milk.
When examined for each sex, TCEP dust concentrations and dust loadings were associated with higher
risk of hay fever/allergies in females (RR: 1.29; 95% CI: 1.16-1.44, RR: 1.33; 95% CI: 1.21-1.47,
respectively), and sex differences were statistically significant for both dust loadings and concentrations
(Pinteraction^ 0.001). Significant associations between TCEP in dust and hay fever/allergies were not seen
when males were examined separately (Mendv et al.. 2024).
Lung Function: Prenatal dust TCEP loadings were associated with lower FEVi at age 5 years (P: -6.17;
95% CI: -11.09, -1.25). In contrast, urinary BCEP at 26 weeks of gestation was associated with a higher
FEVi (P: 4.88; 95% CI: 0.04-9.73), which would seemingly suggest a possible beneficial association
(Mendv et al.. 2024).
Navaranian et al. (2021) used a case-cohort design nested in a Canadian cohort study to investigate the
association between 29 OPEs in dust vacuumed from children's homes (including sleeping quarters)
when they were 3 to 4 months old and childhood asthma at 5 years old or recurrent wheeze at ages 2 to 5
years. For each chemical, odds ratios (ORs) were determined between each of the three higher exposure
Page 270 of 638
-------
quartiles compared with the lowest quartile. Models were adjusted for study site, household income,
child sex, and parental history of asthma. TCEP was not associated with asthma or wheeze. Pearson
correlations showed that TCEP was not significantly associated with other measured OPEs (range: 0.00-
0.13). EPA assigned a medium overall quality determination to this study.
Araki et al. (2020) studied the association between flame retardants and phthalate chemicals of 128
elementary school-aged children in Japan with wheeze and allergic symptoms (i.e., rhinoconjunctivitis,
eczema) as determined by questionnaire; urine samples were taken during the home visit when the
questionnaire was administered. The participation rate was low (only 2.9%).
TCEP in urine (standardized for creatinine content) was not significantly associated with wheeze,
rhinoconjunctivitis, or eczema when comparing the second or third tertile of exposure with the first
tertile. The chemical mixture was associated with increased rhinoconjunctivitis (OR: 2.60; 95% CI:
1.38-5.14). Models were adjusted for sex, grade, household income, and dampness index. Spearman
correlation coefficients between TCEP and other related chemicals were not statistically significant and
ranged from -0.122 and 0.094. Araki et al. (2020) received a high overall quality determination.
Liao et al. (2023) assessed the association between serum TCEP concentrations and Sjogren's syndrome
(SjS) in a study of 138 SjS patients and 145 controls in Hangzhou, China. The only statistically
significant associations found between TCEP serum levels and SjS were for an analysis using exposure
quartiles in which the observed relationship with SjS was non-linear. Compared to the lowest TCEP
exposure quartile (first quartile), the odds of Sj S were significantly lower in the second and third quartile
and higher in the fourth quartile in the crude analyses. After adjusting for potential confounders, the
inverse associations observed for the second and third quartile remained statistically significant, but the
positive association for the fourth quartile was attenuated and no longer statistically significant (p =
0.143). Despite some issues with the terminology used, the methods described in the paper are generally
consistent with a case-control design, which would be appropriate because SjS is a rare disease. Logistic
regression was used, which is appropriate for a case-control study. SjS is substantially more common in
women than men and may be associated with age. Although controls were matched to cases based on
gender during participant selection, the statistical analyses didn't account for this matching, and some
but not all analyses were stratified based on gender. Controls and cases weren't matched by age during
participant selection, but the analyses adjusted for age as well as other potential confounders (smoking
status and drinking habits). There are temporality concerns relevant to interpreting the findings of this
study, particularly the limitation that serum TCEP levels were only assessed at a single timepoint, which
didn't precede the development of the outcome. Due to limitations and limited details regarding
methods, EPA assigned a medium overall quality determination to this study.
Canbaz et al. (2015) did not identify an association between TCEP levels from mattress dust in Swedish
homes where 2-month-old children lived and the subsequent development of asthma when the children
reached ages 4 or 8 years in a medium-quality study.
Laboratory Animals
NTP (1991b) reported no chemical-related changes in hematological parameters in rats or mice after 66
weeks of exposure and no histopathological changes in bone marrow, lymph nodes, spleen, or thymus;
rats did show a statistically significant increased trend in mononuclear cell leukemia with increasing
dose in a species (F344 rats) with a high rate of this cancer. No other in vivo animal toxicity studies were
identified that studied specific immune system changes.
Page 271 of 638
-------
Mechanistic Information
Three in vitro studies examined immune effects. Zhang et al. (2017a) found that 12.5 to 200 mg/L TCEP
was associated with a decrease of roughly 13 to 16 percent in the production of IL-6 at 24 or 48 hours in
the supernatant of human hepatocytes (L02 cells). Another study (Zhang et al.. 2017b) found that TCEP
at 50 and 200 mg/L in liver cells for 24 or 48 hours also resulted in decreased IL-6 protein, up to roughly
33 percent. The adversity of these changes in IL-6 are unclear. For example, IL-6 can act as a pro-
inflammatory cytokine but may also have anti-inflammatory properties (Scheller et al.. 2011). Using the
human hepatocellular carcinoma cell line HepG2, Krivoshiev et al. (2018) found that TCEP altered gene
expression of effector and regulatory proteins in the inflammatory process and concluded that TCEP
may influence inflammation and alter immune function. Zhang et al. (2017b) found that liver cells co-
exposed to both TCEP and benzo[a]pyrene activated pathways associated with inflammation and
increased expression of pro-inflammatory cytokines, whereas exposure to TCEP alone did not yield
similar changes.
Evidence Integration Summary
Evidence from epidemiological studies in Canada, Sweden, or Japan did not identify an association
between TCEP and childhood asthma, but a U.S. study found multiple associations with TCEP and
respiratory outcomes. A study of Sj S in China did not show an effect overall but when investigating
different exposure quartiles, the authors identified a negative association at lower concentrations and a
non-significant positive association at the highest concentration after adjusting for other factors; it is
possible that there is a non-monotonic relationship. Mendv et al. (2024) note that short half-lives of
chemicals such as TCEP (when measuring metabolites in urine) and lack of accounting for dietary and
outdoor concentrations when using dust concentrations and loadings make interpretation of results
somewhat difficult.
Overall, Mendv et al. (2024) is a high-quality study conducted in the United States that identified
multiple positive associations with wheeze, allergy symptoms, and respiratory infections and was more
robust in the numbers of measurements of exposure and outcomes than other studies of wheeze, allergy
symptoms, and asthma. It is not clear why there were differences in outcomes among studies but Araki
et al. (2020) had a very low participation rate and may not reflect a wider group. Also, Canbaz et al.
(2015) focused only on children with diagnosed asthma (not symptoms). Differences in timing of when
exposures were measured may have also led to differences in outcomes. EPA concluded that the human
evidence is slight for immune effects, specifically for wheeze, allergies, and respiratory infections.
Animal studies did not identify histopathological changes in immune-related organs or in hematological
parameters. A statistically significant increased trend in mononuclear cell leukemia with increasing dose
was seen in rats. In mechanistic studies, TCEP was associated with decreases in an inflammatory
cytokine and altered gene expression of inflammatory proteins in two studies, but a third study identified
inflammatory changes only after co-exposure with benzo[a]pyrene. EPA has not determined whether the
decreases of IL-6 in mechanistic studies are adverse.
Based on slight human evidence and indeterminate animal and mechanistic evidence, EPA concludes
that the evidence suggests but is not sufficient to conclude that TCEP exposure causes immunological or
hematological effects in humans under relevant exposure circumstances.
5.2.3.2.5 Thyroid
Humans
From the updated literature search (2019-2024), EPA identified a case-control study (Liu et al.. 2022) in
Shandong Province, eastern China that evaluated risk of thyroid cancer, but also measured the
Page 272 of 638
-------
association between TCEP in serum and thyroid hormone levels normalized by lipid weight within the
control group. Among female controls, increases in TCEP exposure were associated with decreases in
triiodothyronine (T3), free T3, and free thyroxine (T4) (p < 0.05). Thyroid stimulating hormone (TSH)
was increased with increasing TCEP exposure (p < 0.05) among females. In male controls, the only
statistically significant change was a decrease in free T4 (p < 0.05); males showed a decrease in TSH
without statistical significance, and there was no obvious relationship with T3 for males. The model was
adjusted for age, body mass index (BMI), smoking, alcohol consumption, and diabetes status. EPA gave
this study a medium overall quality determination.
Liu et al. (2022) also evaluated TCEP's association with papillary thyroid cancer but found no
statistically significant differences between case and control groups. Alternately, Hoffman et al. (2017)
identified a statistically significant association between TCEP exposure and papillary thyroid cancer in a
high-quality epidemiology study. Section 5.2.4.1 describes the cancer results in more detail.
Animals
Moser et al. (2015) found no changes in serum levels of total thyroxine (T4) and triiodothyronine (T3) in
Long-Evans dams or offspring at PNDs 6 and 22 when dosed up to 90 mg/kg-day. NTP (1991b)
evaluated histopathological changes in the thyroid and parathyroid in the 16-day, 16-week, and 2-year
rat and mouse studies. In the 2-year study, 12 percent of male mice (6 of 50) exhibited follicular cell
hyperplasia at 350 mg/kg-day vs. 6 percent of controls (3 of 60). NTP (1991b) identified increased
incidences of thyroid neoplasms in rats in a 2-year cancer bioassay; the authors concluded that there is
uncertainty regarding an association with TCEP exposure.
Evidence Integration Summary
Based on these data, both human and animal evidence for non-cancer thyroid effects is indeterminate.
EPA did not identify any mechanistic information specific to the thyroid. Overall, the currently available
evidence is inadequate to assess whether TCEP may cause non-cancer thyroid changes in humans under
relevant exposure circumstances.
5.2.3.2.6 Endocrine (Other)
F0 male and female mice exhibited decreased adrenal weights after administration of 700 mg/kg-day
TCEP for 18 weeks (NTP. 1991a).25 Similar effects were not observed in other studies.
Based on indeterminate human and animal evidence and lack of mechanistic support, the currently
available evidence is inadequate to assess whether TCEP may cause endocrine changes other than
thyroid and reproductive hormones in humans.
Evidence related to reproductive hormones is assessed under discussed in Section 5.2.3.1.2 on
reproductive and developmental toxicity endpoints.
5.2.3.2.7 Lung/Respiratory
Humans
Zhu et al. (2022) evaluated the association between natural log transformed BCEP urinary
concentrations for 987 individuals aged 6 to 79 years who were part of the National Health and Nutrition
Examination Survey (years 2011-2012) and five spirometry measures.
25 Kawashima et al. (1983) measured changes in pituitary weights; this study is being translated and will be evaluated for the
risk evaluation.
Page 273 of 638
-------
BCEP was related to a statistically significant decrease in forced vital capacity (FVC) in two models. In
model 1 (adjusted for age, sex, and race), the beta for the natural log of BCEP concentrations associated
with was -43.37 (with 95% CI: -79.79, -6.95;p = 0.02).
In the second model, FVC was decreased by a greater amount: by -79.34 (95% CI: -143.29, -15.39;p =
0.016). This model was adjusted for age, sex, race, body mass index (BMI), serum cotinine, smoking
status, physical activity, family poverty/income ratio, educational level, and urinary creatinine.
FEVi divided by FVC was significantly related to increased BCEP in urine: P of 0.005 (95% CI: 0.001,
0.01; p = 0.007) for the first model. However, the second model was not statistically significant for this
endpoint. The study received a medium overall quality determination.
In a high-quality study described in the above section on immunological and hematological outcomes,
TCEP dust loadings were associated with lower FEVi at age 5 years (P: -6.17, 95% CI: -11.09, -1.25),
but urinary BCEP at 26 weeks of gestation was associated with a higher FEVi (P: 4.88, 95% CI: 0.04-
9.73), which would seemingly suggest a possible beneficial association (Mendv et al.. 2024).
Laboratory Animals
Lung weight changes were identified after 16 weeks (an increase of 17.5 percent in absolute weight in
350 mg/kg-day female rats and decreases of 9 percent in absolute weight at 700 mg/kg-day in female
mice with relative-to-body lung weight decreases of 11.7 and 8.4 percent at 350 and 700 mg/kg/day,
respectively).26 No changes were identified at the 66-week interim sacrifice in the 2-year bioassay, and
no non-cancer changes in histopathology were seen in rats or mice after two years other than increased
hemorrhage with dose in female rats presumed to be associated with cardiovascular collapse in dying
animals (NTP. 1991b). All studies received high overall quality determinations.
Evidence Integration Summary
Based on two epidemiological studies with inconsistent results, human evidence is indeterminate. In
addition, animal data are indeterminate (no relevant histopathological effects, lung weight changes in
studies with high and uninformative overall quality determinations) based on high-quality studies.
Therefore, the currently available evidence is inadequate to assess whether TCEP may cause lung or
respiratory effects in humans under relevant exposure circumstances (see Appendix L.2).
5.2.3.2.8 Body Weight
Humans
Yang et al. (2022). a high-quality study, conducted a prospective cohort study of 340 mother-infant pairs
in Cincinnati, Ohio and examined the association between (1) gestational exposure to the metabolite
BCEP in urine of mothers at 16 and 26 weeks and gestational age; and (2) newborn weight, length,
ponderal index (a relationship between weight and length), and head circumference in offspring.
Yang et al. (2022) found that higher maternal BCEP concentrations in urine at 16 weeks (but not 26
weeks) was associated with a measure of lower birth weight in female infants (P: -0.25 decrease in birth
weight z-score; 95% CI: -0.46 to -0.04) for every 10-fold increase (|ig/L) in BCEP urine concentration.
This study is described in more detail in Section 5.2.3.2.9 below.
Laboratory Animals
26 A decrease was also seen in female rats after 16 days, but the study is uninformative due to a viral infection in the lungs
and salivary glands (NTP. 1991b).
Page 274 of 638
-------
Changes in body weight are of concern and can suggest an underlying toxicity. For TCEP, most studies
ranging from 14 days at doses up to 1,000 mg/kg-day to two years at doses up to 88 and 350 mg/kg-day
in rats and mice, respectively showed no body weight changes greater than 10 percent (Yang et al..
2018a; NTP. 1991a. b). Likewise, dams, fetuses, and pups exhibited no significant body weight changes
when dams were dosed up to 940 mg/kg-day during gestation or gestation and lactation (Moser et al..
2015; Hazleton Laboratories. 1983). Changes were also not observed in adjusted pup weights, F0 or F1
dams at delivery, or in adult males in the continuous breeding study (NTP. 1991a).
Differences in body weights compared with controls were observed in only a few studies. Body weights
of male ICR mice decreased as much as 14.8 percent at 300 mg/kg-day TCEP after 35 days (Chen et al..
2015a). Another study identified a 20 percent increase among female rats after 16 weeks exposure to
350 mg/kg-day TCEP (NTP. 1991b).
In the continuous breeding study, F0 dam weights were decreased at 350 and 700 mg/kg-day from PNDs
7 through 21 (statistically significant trend, with up to 30 percent decrease for the single dam evaluated
at 700 mg/kg-day). In contrast, females in the 350 mg/kg-day group exhibited a 17 percent increase in body
weight at weaning but not during weeks 28 through 30 (NTP. 1991a). Overall, TCEP effects on body weight
were not consistent across studies and when observed, were not consistently increased, or decreased.
Evidence Integration Summary
EPA identified no human studies that had information on body weight changes and therefore, human
evidence is indeterminate. In animal toxicity studies, TCEP effects on body weight were not consistent
across multiple studies. When body weight changes were observed, they were not consistently increased
or decreased. Therefore, the animal data are indeterminate. Overall, the currently available evidence is
inadequate to assess whether TCEP may cause changes in body weight in humans under relevant
exposure circumstances (see Appendix L.2).
5.2.3.2.9 Developmental Toxicity
U.S. EPA (1991) identifies death, structural abnormalities, altered growth, and functional deficits as the
four major manifestations of developmental toxicity. This section describes relevant measurements
related to these outcomes in epidemiolocal studies (as well as pre-term birth and differences in
gestational age of offspring), in prenatal/postnatal studies in mice and rats, and in the continuous
breeding study in mice. This section also describes effects in animals measured during adolescence, a
relevant developmental lifestage (U.S. EPA. 1991). Mating and fertility outcomes resulting from the
continuous breeding study are described in Section 5.2.3.1.2.
Humans
EPA identified four epidemiological studies evaluating growth and gestational age since publication of
the draft risk evaluation. Crawford et al. (2020) was identified in the updated literature search (2019-
2024) conducted by EPA. Peer reviewers identified three additional studies (Oh et al.. 2024; Hernandez-
Castro et al.. 2023b; Yang et al.. 2022). All studies examined the association between BCEP (a TCEP
metabolite) in pregnant mothers' urine and growth/gestational age. The studies are described based on
the size of the cohort, with the largest cohort discussed first. Additional studies related maternal
exposures during gestation are described in Sections 5.2.3.1.1 and 5.2.3.2.4 for and neurotoxicity and
immunological effects, respectively.
Oh et al. (2024). to which EPA assigned a medium overall quality determination, was the largest cohort
study that investigated the association between urinary concentrations of BCEP (identified as BCETP in
the study) and various measures of gestational age and birth weight. BCEP measurements were obtained
Page 275 of 638
-------
primarily from second and third trimesters from 6,646 pregnant women in 16 cohorts across the United
States. BCEP concentrations were divided into non-detect, low exposure, and high exposure categories
for analysis.
Oh et al. (2024) found that the association of the high vs. the non-detect BCEP group with gestational
age and preterm birth differed by sex (p < 0.01). Female offspring had a shorter gestational age (P:
-0:13 weeks; 95% CI: -0.24, -0.03) and a higher odds of pre-term birth for gestational age (odds ratio
[OR]: 1 .27; 95% CI: 1.03, 1.58). The associations for male offspring were not statistically significant but
did show that they had longer gestational age (P = 0.06 weeks; 95% CI: -0.06, 0.18) and lower odds of
pre-term birth (OR: 0.77; 95% CI: 0.55, 1.06). The regression models were adjusted for maternal
measures of race/ethnicity, age at delivery, education, marital status, pre-pregnancy body mass index
(BMI), smoking during pregnancy, and parity, as well as the year and season of sample collection.
The Spearman coefficients (-0.017 to 0.189) showed weak correlations among BCEP and other OPE
metabolite biomarkers, and when regression models were run with all OPE biomarkers, results were
similar to the primary results. The authors noted that removing one cohort at a time demonstrated that
the model was robust; although some estimates were strengthened or weakened with removal of
individual cohorts, the direction of the associations did not change (Oh et al.. 2024).
Among measures of birth weight, high BCEP was associated with lower odds of being small for
gestational age (SGA) for both sexes combined (OR: 0.83; 95% CI: 0.71, 0.96, p < 0.01). In addition to
the model adjustments identified above, this full model (with both sexes combined) also adjusted for
child's sex. In sex-specific models, the association for SGA was statistically significant for males for
both the low BCEP group (OR: 0.73; 95% CI: 0.55, 0.97) and the high group (OR: 0.77; 95% CI: 0.62,
0.97) compared with the non-detects (Oh et al.. 2024).27
Hernandez-Castro et al. (2023b) recruited pregnant women from health clinics, a private obstetrics and
gynecological practice, community meetings, and advertisements in Los Angeles, California for a
prospective cohort study that investigated the association between BCEP and both gestational age and
birthweight among infants. The authors measured BCEP in third trimester urine samples from 421
mothers, who were primarily low income and Hispanic/Latina. The authors also investigated
associations between exposure to mixtures of OPEs and these same birth outcomes. The authors did not
identify any associations for BCEP measurements for the full model or when evaluating effects in
female and male infants separately. The model was adjusted for recruitment site, maternal age, season of
sample collection, gestational age, race/ethnicity, pre-pregnancy BMI, income, education, infant birth
order, maternal hypertensive disorders of pregnancy; infant sex was also used as a covariate for the
gestational age model that included both sexes. When evaluating the OPE mixture, Hernandez-Castro et
al. (2023b) found an association with lower gestational age, primarily based on female infant data but
the association was not statistically significant. Spearman correlations between BCEP and other OPE
metabolite concentrations were generally weak and ranged from 0.02 to 0.22. EPA assigned a medium
overall quality determination to this study.
Yang et al. (2022) conducted a prospective cohort study of 340 mother-infant pairs in Cincinnati, Ohio
and examined the association between (1) gestational exposure to the metabolite BCEP in urine of
mothers at 16 and 26 weeks and gestational age; and (2) newborn weight, length, ponderal index (a
relationship between weight and length), and head circumference in offspring. BCEP at 26 weeks was
27 Table S8 identifies 0.73 and 0.77 as (3 values. However, the text describes these male-specific effects as the odds of
occurrence and Table 4 presents ORs for SGA (small for gestational age). Therefore, EPA assumes that the identification of
these numbers in Table S8 should have been as ORs instead of (3 values.
Page 276 of 638
-------
positively associated with gestational age for both sexes combined (P: 0.33-week increase; 95% CI:
0.06-0.61) for every 10-fold unit increase in BCEP urine concentration (in |ig/L), but not when stratified
by sex. Increases of logio BCEP concentrations at 26 weeks were also associated with lower pre-term
birth (RR = 0.48; 95% CI: 0.27, 0.83), but the authors suggested that the results related to pre-term birth
need to be interpreted with caution because there were only 30 pre-term births. The model was adjusted
for maternal age at delivery, race, household income, education, marital status, infant sex, parity, and
pre-pregnancy BMI, serum cotinine/blood lead levels at 16 weeks gestation.
Yang et al. (2022) ran two other models and identified similar associations with pre-term birth or
gestational age. First, the Cox proportional hazard model showed that a logio increase in BCEP
concentrations at 26-weeks was inversely associated with pre-term birth (hazard ratio (HR): 0.40; 95%
CI: 0.19, 0.83). A similar relationship using the Cox model was seen for the average of the 16 and 26-
week BCEP concentrations and pre-term birth (HR: 0.41; 95% CI: 0.17, 0.98). When divided into
tertiles with the lowest exposure tertile as reference, the 26-week BCEP concentration was positively
associated with gestational age (p for trend = 0.02). These models were adjusted for the same covariates
identified for the above model (Yang et al.. 2022).
Yang et al. (2022) also found that higher maternal BCEP exposure at 16 weeks (but not 26 weeks) was
associated with lower birth weight z-scores28 in female infants (P: -0.25 decrease in birth weight z-score;
95% CI: -0.46 to -0.04) for every 10-fold increase (|ig/L) in BCEP urine concentration. The 10-fold
(|ig/L) BCEP urine concentration increases were associated with decreased female infant length z-
scores, with p of-0.31 (95% CI: -0.56, -0.07) at 16 weeks and p of-0.18 (95% CI: -0.35, -0.02) at 26
weeks. EPA assigned a high overall quality determination to this study.
A pilot prospective cohort study (Crawford et al.. 2020) evaluated BCEP (identified as BCHP in the
study) in urine among 56 pregnant women representing the general population of Rhode Island and
examined the association between BCEP and various anthropometric measures in children born to these
mothers at birth and 6 weeks of age. Seventy-one percent of the women provided samples for all three
trimesters. BCEP was associated with an overall increased thigh skinfold thickness in males and females
(P: 0.34 mm; 95% CI: 0.16, 0.52); and increased subscapular skinfold thickness in males (P: 0.14 mm;
95% CI: 0.002, 0.28) (p = 0.05 for the effect modification by infant sex). Results were adjusted for
maternal age at delivery, income, pre-pregnancy BMI, parity, infant sex, and age at the time of
measurement. As noted by Crawford et al. (2020). skinfold thickness measurements correlate well with
subcutaneous fat distribution. Although researchers may measure skinfold thickness differently from
each other, they received the same training, and any measurement errors were expected to be minimized.
BCEP was not associated with infant weight, length, or head and abdominal circumferences. Also, there
were no statistically significant effects of BCEP on infant feeding behavior, but BCEP did show a
tendency for increased general appetite (P: 0.11; 95% CI: -0.04, 0.27). Crawford et al. (2020) note that
based on the number of pregnant women included in this pilot study, it was underpowered to investigate
the effects of OPE mixtures. The authors also did not analyze correlations among chemical exposures.
EPA gave the study a high overall quality determination.
Percy et al. (2021). Percy et al. (2022). and Hernandez-Castro et al. (2023a) examined prenatal BCEP
concentrations and neurobehavioral measures and found some negative association with IQ for
individuals in certain groups associated with lower SES in one study (Percy et al.. 2022) (see Section
5.2.3.1.1).
28 Z-scores relate infant size measures to the population distribution. For example, an infant with an average body weight has
a weight z-score of 0 and an infant with a weight one standard deviation higher than the mean population weight has a z-
score of +1.
Page 277 of 638
-------
Mendv et al. (2024) identified some associations between mothers' exposure to BCEP during gestation
and children's respiratory symptoms as described in Section 5.2.3.2.4.
Page 278 of 638
-------
Table 5-46. Associations between BCEP (TCEP Metabolite) in Urine of Pregnant Women and Growth and Gestational Age
#Mom:
Infant
Pairs
Geographic
Region
BCEP During
Pregnancy"
Birth Outcomes
Measured6
Statistically Significant
Associations
Model Adjustments
Co-Exposures
Measured
Citation,
OQD
6,646
16 individual
cohorts
across the
United States
Second and third
trimesters
exposure groups:
non-detect, low, high
[50%ile: 0.52 ng/mL;
5%ile < LOD (0.02);
95%ile: 8.22]
GA, pre-term, early
term, full term
post/late term
BW at birth (BW-GA
z-scores, SGA, LGA,
LBW)
Both sexes: vp SGA (high
BCEP)
Females: vp GA,
1s pre-term0
Males: vp SGA (low and
high BCEP)
Maternal age at delivery,
education, race/ethnicity,
marital status, pre-pregnancy
BMI, smoking during
pregnancy, parity, as year
and season of sample
collection
Weak correlations
between BCEP and
other OPEs (-0.017
to 0.189); multiple
OPE model
supports BCEP
results
Oh et al.
(2024)
OQD =
Medium
421
Los Angeles,
California;
primarily
low-income
Hispanic/
Latina
Third trimester (31.5
+/- 2.0 wks)
[GM = 0.31 ng/mL;
25, 50, 75 %ile =
0.03,0.53, 1.62; min
- ND; max -168;
LOD - 0.02]
GA, BW (as BW-GA
z-scores)
BCEP: None for full model
or by sex
OPE mixture, females
(primarily): vPGA
Recruitment site, maternal
age, season of sample
collection, gestational age,
race/ethnicity, pre-pregnancy
BMI, income, education,
infant birth order, maternal
hypertensive disorders of
pregnancy; infant sex (GA
model with both sexes)
Weak correlations
between BCEP and
other OPEs (0.02
to 0.22); multiple
OPEs not included
in models
Hernandez-
Castro et al.
(2023b)
OQD =
Medium
340
Cincinnati,
Ohio
>50% non-
Hispanic
16 and 26 wks
[GM (GSD): 0.60
(3.16) ng/L at 16 wks
0.51 (4.33) ng/Lat
26 wks]
GA, pre-term at birth
newborn BW, length,
ponderal index
(relationship between
weight and height),
and head
circumference
Both sexes: 1s GA (main
model and tertiles/trend test,
26-wk BCEP); vp pre-term
birth (main model and Cox,
26 or avg. 16/26-wk BCEP)
Females:
16-wk BCEP:
vp BW z-score;
ave 16/26 wk BCEP:
^ length z-score
Maternal age at delivery,
race, household income,
education, marital status,
parity, and pre-pregnancy
BMI, serum cotinine/blood
lead levels at 16 weeks
gestation; infant sex in
models with both sexes
No
Yans et al.
(2022)
OQD =
Medium
56
Rhode Island
12, 28, and/or 35
wksrf
[Median = 0.31 ug/L;
0.17 to 0.60 for
interquartile range]
Skinfold thickness,
BW, length, head,
and abdominal
circumferences,
feeding behavior at
birth and 6 wks
Both sexes: 1s thigh skinfold
thickness
Males: Is subscapular
skinfold thickness
Maternal age at delivery,
income, pre-pregnancy BMI,
parity, and age at the time of
measurement; infant sex in
models with both sexes
No
Crawford et
al. (2020)
OQD = High
" BCEP concentrations were normalized by specific gravity of the urine. h GA = Gestational age; BW = body weight; SGA and LGA (<10th and >90th percentiles of BW at
gestation, respectively); LBW (birthweight <2,500 g at >37 weeks gestation);c Males had t^GA and ^ pre-term birth that was not statistically significant;d 71 percent of mothers
had data for all three timepoints.
Page 279 of 638
-------
Laboratory Animals
EPA identified three prenatal/postnatal animal studies. Two studies received high overall quality
determinations. Hazleton Laboratories (1983) administered 940 mg/kg-day TCEP via oral gavage to
female CD-I mice from GD 7 to 14. Dams exhibited clinical signs of neurotoxicity but no differences in
measures of live or dead pups per litter. In addition, there were no changes in fetal or pup weights.
Similarly, Long-Evans rat dams were dosed from GD 10 to PND 22 via oral gavage at 0, 12, 40, and 90
mg/kg-day (decreased from 125 mg/kg-day after 5 days) in the developmental neurotoxicity study
described in Section 5.2.3.1.1. There were no differences in litter size on PND 2 or changes in offspring
weight (Moser et al.. 20 1 5).29-30
Kawashima et al. (1983) evaluated effects of TCEP exposure on developmental outcomes after dosing
pregnant Wistar rats via oral gavage with 0, 50, 100, or 200 mg/kg-day from GD 7 through 15, and EPA
assigned it a medium overall quality determination. Seven of 30 dams (23%) dosed with 200 mg/kg-day
died during gestation, as noted in Section 5.2.3.1.1. Kawashima et al. (1983) is the only study that
evaluated teratogenicity and skeletal variations after TCEP exposure. The authors identified no
increased malformations or variations. Offspring exhibited a 27 percent decrease in absolute pituitary
weight at the highest dose (with maternal toxicity) but no changes in multiple other organ weights
(Kawashima et al.. 1983). Offspring exposed to TCEP in utero showed no significant differences from
controls in numbers of surviving pups, sex ratio, body weight, or mortality rate. Section 5.2.3.1.1
describes the functional and neurobehavioral effects in male offspring from this study.
In the RACB protocol NTP (1991a). the 350 and 700 mg/kg-day mice exhibited decreases in average
number of litters per pair and live pups per litter (p < 0.001).
During crossbreeding of F0 mice, the 700 mg/kg-day male x control female group yielded decreased
live F1 pups per litter (statistical analysis not possible because only one litter was delivered). Results of
700 mg/kg-day females crossed with control males also led to decreases in live F1 pups per litter (p <
0.01 males; p < 0.05 both sexes). Outcomes from treated males x control females were more
pronounced, with production of just 1 litter with 3 live pups vs. 12 litters and 7.2 live pups per litter
from treated females x untreated males. The control x control group resulted in 12 litters and 10.3 live
pups per litter compared with either 700 mg/kg-day males or females crossbred with controls (NTP.
1991a).31-32
After F1 breeding, there were decreased numbers of live F2 pups per litter at the highest dose of 350
mg/kg-day (p < 0.05). Although live male F2 pups per litter were also reported as being significantly
decreased at 175 mg/kg-day (NTP. 1991a). EPA identified a discrepancy in NTP's Table 4-4 in the
proportion of males.
Effects were more pronounced across generations. The same dose (e.g., 350 mg/kg-day) resulted in
fewer live F2 pups per litter (7.6) than live F1 pups per litter (10.1) (NTP. 1991a).
29 Limited information from the unavailable Russia inhalation study in rats, Shepel'skaia and Dvshginevich (1981) identified
decreased body weight and crown rump length in rat offspring at 0.5 mg/m3.
311 NTP (1991a) identified no effects on sex ratio in the first generation, and although significant differences in sex ratio from
controls were observed in the second generation, there is uncertainty in the change due to a discrepancy in reporting of
proportion of male offspring born alive at the highest dose (0.41 vs. 0.45). Moser et al. (2015) did not identify effects on sex
ratio. Hazleton Laboratories (1983) did not describe whether sex ratio was measured.
31 The number of breeding pairs examined ranged from 18 to 20 among dose groups.
32 Shepel'skaia and Dvshginevich (1981) cited in (NTP. 1991a) (unobtainable Russian abstract) resulted in dams with
significantly decreased litter size and increased pre- and post-implantation loss at 1.5 mg/m3.
Page 280 of 638
-------
Mechanistic Information
Yonemoto et al. (1997) identified an IP50 (inhibitory concentration for cell proliferation) 3,600 |iM of
TCEP using rat embryo limb bud cells. The ID50 (inhibitory concentration for differentiation) was
identified as 1,570 |iM. The authors concluded that the high proliferation to differentiation ratio
suggested that TCEP should be investigated more fully for developmental toxicity.
In vivo and in vitro studies found TCEP to affect male reproductive hormones as noted in Section
5.2.3.1.2 including decreases in both testosterone secretion and decreases in a gene associated with
testosterone synthesis in mouse Leydig (TM3) cells (Chen et al.. 2015a; 2015b). These reproductive
studies may support observed developmental effects based on effects on offspring viability observed
after crossbreeding treated males with control females.
In other in vitro studies, TCEP was not associated with estrogenic or anti-estrogenic effects or changes
in AR-mediated gene expression or ERa and AhR target gene activation (Reers et al.. 2016; Follmann
and Wober. 2006). TCEP did not exhibit estrogenic activity in in MCF-7 cells but did yield anti-
estrogenic activity when co-treated with E2 (Krivoshiev et al.. 2016).
Evidence Integration Summary
EPA located four human epidemiological studies that found changes in gestational age and some growth
measures, including some increased skinfold thickness in offspring associated with maternal urine
concentrations of BCEP, a TCEP metabolite, but did not identify any associations with other body
composition measurements. One study identified a decrease in IQ among lower SES children and another
identified increased respiratory effects among children whose mothers' showed exposure to BCEP during
gestation (or at delivery).
EPA has concluded that the human evidence is slight for developmental effects given that there are
associations for several endpoints related to growth, gestational age, and other effects, but there are some
inconsistencies among results (e.g., increased gestational age for both sexes in one experiment; decreased
gestational age and increased pre-term birth for females in another) and a lack of effect on gestational age
and growth in one of the larger studies Hernandez-Castro et al. (2023b).
Animal studies show slight evidence for developmental effects. Developmental outcomes such as
decreased live pups per litter were observed in the NTP RACB study (described in Section 5.2.3.1.2)
with increased severity in the second generation. However, the prenatal and prenatal/postnatal studies
did not result in developmental outcomes, except under severe maternal toxicity. Although differences
in study protocols between the RACB and prenatal studies may explain differences in outcomes, there is
only one acceptable study that investigated exposure prior to gestation. There is some support for the
potential for developmental effects based on male reproductive toxicity observed in animal studies (see
Section 5.2.3.1.2).
The limited mechanistic evidence of reproductive toxicity can be relevant as considerations for
developmental toxicity. EPA considers the supporting mechanistic data to be slight.
Overall, EPA concluded that evidence suggests but is not sufficient to conclude that TCEP exposure
causes developmental toxicity in humans under relevant exposure circumstances. This conclusion is
based on effects in four epidemiological studies using the TCEP metabolite BCEP as a biomarker that
Page 281 of 638
-------
are inconsistent and fertility-related changes in the RACB study. The oral studies in mice and rats used
as a basis for this decision evaluated doses of 12 to 700 mg/kg-day (Table Apx L-3).
5.2.4 Cancer Hazard Identification, MOA Analysis, and Evidence Integration
The sections below outline human (Section 5.2.4.1) and animal evidence (Section 5.2.4.2) for
carcinogenicity as well as and an MOA summary (Section 5.2.4.3) and a summary of evidence
integration conclusions (Section 5.2.4.4).
5.2.4.1 Human Evidence
One high-quality case-control cancer study examined the association between TCEP/other flame-
retardant exposure and papillary thyroid cancer in adults living near Duke University in North Carolina
(Hoffman et al.. 2017). TCEP concentrations in dust were measured in 70 age- and gender-matched
cases and controls in 2014 to 2016; no biological measurements were collected for TCEP. The authors
identified a median TCEP concentration of 400 ng/g in dust. Diagnosis of papillary thyroid cancer was
positively associated with TCEP concentrations above the median. The odds ratio is 2.42 (CI: 1.10-
5.33) (p< 0.05).
In contrast, another case-control study with a medium overall quality determination that was conducted
in Shandong Province, eastern China (Liu et al.. 2022) did not identify any statistically significant
associations between TCEP and papillary thyroid cancer. The study compared upper quartiles of TCEP
concentration in serum normalized by lipid weight with the lowest quartile concentration. The results
were adjusted for age, BMI, smoking, alcohol consumption, and diabetes status in the regression model.
ORs were greater than one only for males (1.24) for the 50 to 75th percentile (but not the highest (>75th)
percentile) and only for females for the highest percentile comparison (1.14). None of the associations
were statistically significant. See Section 5.2.3.2 for information on TCEP's association with thyroid
hormones from this study.
Li et al. (2020) examined the association between TCEP and other OPEs in plasma and prevalence of
gastrointestinal and colorectal cancers (all stages) in Wuhan, China. There were 34 cases of
gastrointestinal cancer and 40 cases of colorectal cancer and 62 controls, who were health individuals
without cancer. EPA gave this study a medium overall quality determination. TCEP was detected in
gastrointestinal and in colorectal cancer patients more frequently than in the control group, for which
TCEP was not detected (p < 0.01 for both comparisons); the concentrations were also higher in each of
the cancer groups (p < 0.01). However, there was no statistically significant association between TCEP
concentrations in plasma and the presence of either cancer when using binary logistic regression for an
unadjusted model or for a model adjusted for sex and gender.
Liu et al. (2021) investigated TCEP in plasma and female-related cancers. A total of 258 women were
recruited from a hospital in Wuhan, China in April 2019 with benign breast tumors (n = 45), breast
cancer (n = 73), benign uterine tumors (n = 62), and cervical cancer (n = 78). The concentration of
TCEP in the benign uterine tumor group was higher than the cervical cancer group (p < 0.05). No other
associations were identified for TCEP between the benign and cancerous tumor groups. EPA gave this
study a low overall quality determination partly based on lack of controls without tumors as well as
small sample sizes per group, and use of cross-sectional design to evaluate a chronic risk such as cancer.
5.2.4.2 Animal Evidence
EPA identified one oral NTP cancer bioassay in which F344/N rats B6C3Fi mice (50 per sex per dose of
each species) were administered TCEP in corn oil via oral gavage for 5 days per week for 104 weeks.
Page 282 of 638
-------
Rats received 0, 44, or 88 mg/kg and mice received 0, 175, or 350 mg/kg (NTP. 1991b). The study
received high overall quality determinations for the tumor incidence data.
NTP (1991b) identified multiple tumors and concluded that there is clear evidence of carcinogenic
activity of renal tubule adenomas in male and female rats. The authors also concluded that thyroid
follicular cell neoplasms and mononuclear cell leukemia in rats may have been related to TCEP
administration but acknowledge uncertainty related to this association. There was equivocal
carcinogenic evidence based on marginally increased incidence of renal tubule cell neoplasms in for
male mice and marginally increased incidence of harderian gland adenomas in female mice.33
Kidney Tumors
Rats: At the 66-week sacrifice, one high-dose male had a renal tubule adenoma. At the end of the study,
high-dose male rats exhibited increased incidences of renal tubule adenomas (48%) vs. control rats (2%)
(p < 0.001) and a dose-response trend was evident (p < 0.001). Male rats also exhibited hyperplasia of
the renal tubule epithelium, with 48 percent incidence at the high dose (vs. 0 percent in controls). One
control and one high dose male developed a renal tubule carcinoma. High-dose females had a lower
incidence of renal tubule adenomas (10%), but incidence was higher than controls (0%) (p < 0.05) with
a statistically significant dose-response trend (p < 0.001). High dose females also exhibited a 32 percent
incidence of focal hyperplasia of the renal tubule epithelium vs. 0 percent in controls.
Rats exhibited lower survival rates at 88 mg/kg-day after dosing with TCEP: 51 vs. 78 percent in
controls in males and 37 vs. 66 percent in controls for females. Female survival started to decrease at
week 70 and many rats exhibited brain lesions, whereas males' decreased survival was limited to the
final month of the study.
Mice: Mice exhibited no decreases in survival. At the end of the study, eight percent of high-dose male
mice had either renal tubule adenomas or adenocarcinomas compared with 2 percent in controls. Only
one low dose female exhibited a renal tubule adenoma. Six percent of mice exhibited renal tubule cell
hyperplasia. All treated mice had statistically significant increases in enlarged nuclei in renal tubule
epithelial cells (NTP. 1991b). No kidney-related lesions were observed at the 66-week interim
sacrifice.34
Other Tumors
Hematopoietic System: Mononuclear cell leukemia (MNCL) was increased in male rats at both doses
(28 and 26 percent, respectively) vs. 10 percent in controls. Because these are fatal neoplasms, life table
analyses are considered important and showed statistical significance for the low and high doses vs.
controls (p < 0.05) and for a dose-response trend (p = 0.01). Female rats exhibited a slight increase at the
high dose (40%) compared with controls (28%) and exhibited a dose-response trend (p < 0.01).
Although MNCL may relate to TCEP exposure, the increase in male rats was not clearly dose-related
and was partly due to incidence that was lower than expected in the controls. In addition, historical
33 Takada et al. (1989) dosed ddY mice at 0, 0.012, 0.06, 0.3, or 1.5 percent TCEP to ddY mice in the diet for 18 months and
identified increased incidence of tumors in multiple target organs; this study is not in English and was not translated or
evaluated for data quality. Takada et al. (1989) was, however, described in the 2009 PPRTV for TCEP (U.S. EPA. 2009).
U.S. EPA (2009) presented estimated doses for this study as 0, 9.3, 46.6, 232.8, and 1,687.5 for males and 0, 10.7, 53.3,
266.7, and 1,875 for females using measured data for body weight and food consumption from the bioassay in the following
equation: %diet x 10,000 x estimated food consumption)/estimated body weight.
34 Takada et al. (1989) identified an incidence of 82 percent renal cell adenomas and carcinomas in male mice at the highest
concentration vs. 4 percent in controls (p < 0.01).
Page 283 of 638
-------
control values for these neoplasms are variable and all incidences in the current study were within
historical controls (NTP. 1991b).35
Thyroid: Other notable tumors in rats identified in the NTP (1991b) bioassay included slightly increased
incidences of thyroid combined follicular cell adenomas and carcinomas observed in high-dose males
(10 vs. 2 percent control males) and in high-dose females (8 vs. 0 percent in controls). The incidence in
females exhibited a statistically significant dose-response trend and pairwise comparison at the highest
dose (p < 0.05). NTP concluded that these tumors may be related to TCEP exposure. However, the
increases were considered marginal. In addition, female rats did not exhibit thyroid follicular
hyperplasia, and NTP (1991b) states that most thyroid carcinogens also cause hyperplasia.
Hardericm Gland: At the 66-week sacrifice in NTP (1991b). two high-dose female mice had adenomas
of the harderian gland and a third had a harderian gland carcinoma. In female mice, combined incidence
of harderian gland adenomas and carcinomas from both the 66-week and terminal sacrifices were
increased (5, 13, and 17 percent for controls, low, and high doses). Both the high-dose incidence vs.
controls and dose-response trend were statistically significant (p < 0.05).36
Liver: Male mice exhibited a significant positive trend for hepatocellular adenoma (p < 0.05) with 40,
36, and 56 percent incidence in controls, 175, and 350 mg/kg-day, respectively. However, the increase at
the high dose compared with controls was not statistically significant and there was no increase in
hepatocellular carcinomas compared with controls. Male mice also exhibited increased eosinophilic foci
(16 vs. 0 percent at the high dose compared with controls) but no increase in basophilic or clear cell foci,
which constitutes a morphological continuum with hepatocellular adenoma (NTP. 1991b).37
Uterine: Three female rats had uterine stromal sarcomas at the high dose but none in controls or the low-
dose group. Although the trend test was significant (p < 0.05), the incidence in the high-dose group was
not significantly greater than in concurrent or historical controls and thus, NTP (1991b) concluded that
the uterine tumors were not related to TCEP administration.
Mammary Gland: Three high-dose female mice had adenocarcinomas of the mammary gland with a
positive trend (p < 0.05). However, a fibroadenoma occurred in a female control; there was no
significant trend for fibroadenoma, or adenocarcinoma combined; and the incidence of adenocarcinomas
is within female historical vehicle controls. Therefore, NTP (1991b) concluded that the mammary gland
adenocarcinomas were not related to TCEP treatment.
5.2.4.3 MOA Summary
The U.S. EPA (2005b) Guidelines for Carcinogen Risk Assessment defines mode of action as "a
sequence of key events and processes, starting with the interaction of an agent with a cell, proceeding
through operational and anatomical changes and resulting in cancer formation." Hard (2018) has
identified modes of action for renal tubule carcinogens that include direct DNA reactivity, indirect DNA
reactivity resulting from formation of free radicals, bioactivation involving glutathione conjugation,
mitotic disruption, sustained cell proliferation resulting from direct cytotoxicity, sustained cell
proliferation after disruption of a physiologic process (such as alpha 2u-globulin nephropathy), chemical
exacerbation of chronic progressive nephropathy among others.
35 Takada et al. (1989) found increased incidence of leukemia (type not specified) in female ddY mice (18 percent at ~ 266.7
and 1,875 mg/kg-day) compared with two percent in controls (p < 0.05).
36 There were no increases in harderian gland tumors in male or female ddY mice (Takada et al.. 1989).
37 Takada et al. (1989) identified increased hepatocellular adenomas or carcinomas in male ddY mice of 26 and 38 percent at
232.8 and 1,688 mg/kg-day in the diet compared with 8 percent in controls (p < 0.01).
Page 284 of 638
-------
The target organ with the most robust evidence of carcinogenicity for TCEP is the kidney. In addition to
genotoxicity information on multiple cell types, EPA summarizes other biochemical and cellular effects
primarily in renal cells and kidneys. EPA did not conduct a formal analysis using concordance tables to
separately evaluate postulated MO As according to the International Programme on Chemical Safety
(IPCS) Conceptual Framework for Evaluating a Mode of Action for Chemical Carcinogenesis (Sonich-
Mullin et al.. 2001). Available data from in vitro studies identified effects associated with TCEP and that
identify a variety of biochemical changes that might be relevant to induction of kidney tumors resulting
from TCEP exposure. However, only sparse in vivo evidence was available to understand the
temporality of precursor events associated with inducing kidney tumors.
Based on extensive data on tests of mutagenicity, EPA concludes that a mutagenic mode of action is not
a likely MOA for TCEP, as noted in Section 5.2.3.2.9 and Appendix M.
TCEP was associated with effects in 28-day studies in kidneys (OSOM and cortex) at 350 mg/kg-day
that included cell cycle deregulation, apoptosis, increases in regenerating tubules, and increased markers
of cell proliferation (but no accompanying proliferative lesions) (2012b; Taniai et al.. 2012a). The
authors surmise that cell proliferation along with aberrant regulation of the cell cycle (e.g., from the G2
phase during which macromolecules are produced to prepare for cell division and through the M phase
of mitosis) may lead to chromosome instability linked to cancer. The accompanying apoptosis may
reflect aberrant cell cycle regulation (Taniai et al.. 2012b). It is also possible that DNA damage may
have been a precipitating factor in the increase of one of the markers (topoisomerase Ila) (Taniai et al..
2012a).
In vitro studies showed that primary rabbit renal proximal tubule cells (PTCs) exposed to TCEP
exhibited altered expression of cell cycle regulatory proteins, reduced DNA synthesis, inhibition of ion-
and non-ion-transport functions (e.g., decreased uptake of sodium, calcium, etc.), and induced
cytotoxicity. Increased expression of pro-apoptotic regulatory proteins and decreased expression of
proteins that inhibit apoptosis were also observed (Ren et al.. 2012; Ren et al.. 2009. 2008).
Studies of other tissues and cell types exposed to TCEP identified cell cycle changes, perturbation of
cell signaling pathways, markers of oxidative stress, impaired mitochondrial function, inhibition of
glutathione, and other effects (see Table Apx L-6).
In NTP (1991b). the authors reported no hyperplasia in rats at the 66-week interim sacrifice in the
narrative (data tables not included). Although focal hyperplasia was observed and can be expected to be
a precursor to tumors, the only related finding regarding kidney tumors at the 66-week sacrifice was a
single renal tubule adenoma seen in female rats. Therefore, evidence of temporal progression from
hyperplasia to adenoma and then carcinoma is not available. At two years, hyperplasia was observed in
male rats, but incidence was slightly lower (0, 2, and 24) than adenomas (1,5, and 24) compared with
hyperplasia at 0, 44, and 88 mg/kg-day. The lack of temporality and limited information on precursor
lesions and their relationship with tumors leads to uncertainty regarding dose-response progression from
hyperplasia to adenomas and carcinomas in males. Female rats did have higher rates of hyperplasia (0,
3, 16) than adenomas (0, 2, 5), at 0, 44, and 88 mg/kg-day, respectively.
Conclusion
Several studies have investigated biochemical and cellular changes in kidneys or renal cells that may be
associated with steps in an MOA for kidney cancer. EPA has not performed a formal analysis on
postulated MO As (e.g., as in Sonich-Mullin et al. (2001)). However, available in vitro studies and a few
Page 285 of 638
-------
in vivo studies that identify multiple biochemical changes that might be relevant to induction of kidney
tumors. There is sparse information on temporality and dose-response of potential pre-cursor events
within the in vivo studies and no clear NOAEL regarding tumor response to be able to confidently model
tumor incidence with a non-linear/threshold dose response analysis as an alternate possible dose-
response.
U.S. EPA's PPRTV (U.S. EPA. 2009) concluded that the overall weight of evidence for mutagenicity is
negative and that no mechanistic data identify specific potential key events in an MOA for kidney or
other tumors induced by TCEP exposure other than a general association with known proliferative and
preneoplastic lesions.
5.2.4.4 Evidence Integration Summary
EPA concludes that TCEP is likely to be carcinogenic to humans using guidance from the Agency's
Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005b). This conclusion is based on clear
evidence of carcinogenic activity in rats based on renal tubule adenomas, equivocal evidence of kidney
tumors in mice, the rarity of the kidney tumors in rodents, and equivocal evidence of several other
tumors in rats or mice. Tumor incidence data are based on oral chronic bioassays in rats and mice that
assessed dose levels between 44 and 350 mg/kg-day. Table_Apx L-6 provides details regarding EPA's
evidence integration conclusion for cancer.38
There is indeterminate evidence in humans from four studies. One identified an association between
TCEP and papillary thyroid cancer (Hoffman et al.. 2017). A second one did not find an association for
the same type of cancer (Liu et al.. 2022). Although TCEP exposure was higher for cases, female-related
cancers (Liu et al.. 2021) and gastrointestinal/colorectal cancers (Li et al.. 2020) were not clearly
associated with TCEP exposure.
In laboratory animal studies, there is evidence of carcinogenicity in two species and both sexes in a
single high-quality study. Evidence for kidney tumors is robust based on increased incidence of renal
tubule adenomas in male and female F344/N rats and marginal increases in these tumors in male
B6C3F1 mice (NTP. 1991b). The rarity of these tumors in F344/N rats and B6C3F1 mice strengthens the
evidence.
Lesions observed in kidneys include focal hyperplasia, renal tubular cell enlargement (karyomegaly),
and adenomas and carcinoma in rats and/or mice (NTP. 1991b). This continuum of has been observed
with renal tubular cell cancer in humans (Beckwith. 1999). Two-year cancer bioassay for a similar
chemical, tris (2,3-dibromopropyl) phosphate (CASRN 126-72-7), also resulted in kidney tumors in
male and female rats and male mice and karyomegaly in mice (NTP. 1991b).
For MNCL, evidence is slight. NTP (1991b) observed significant pairwise increases and dose-response
trends of MNCL in male and female F344/N rats. However, MNCL is common in F344 rats, its
spontaneous incidence varies widely, and incidences in male rats exposed to TCEP were within
historical controls. Occurrence of these tumors is rare in mice and other strains of rats (Thomas et al..
2007). Further, there is uncertainty regarding similarity to tumors in humans. MNCL may be similar to
large granular lymphocytic leukemia (LGLL) in humans (Caldwell et al.. 1999; Caldwell. 1999;
Reynolds and Foon. 1984). particularly an aggressive form of CD3- LGL leukemia known as aggressive
natural killer cell leukemia (ANKCL) (Thomas et al.. 2007). However, Maronpot et al. (2016) note that
38 Using the 2021 Draft Systematic Review Protocol (U.S. EPA. 2021a). the equivalent conclusion is that TCEP likely causes
cancer in humans under relevant exposure circumstances.
Page 286 of 638
-------
ANKCL is extremely rare with less than 98 cases reported worldwide, and the authors contend that
ANKCL has an etiology related to infection with Epstein-Barr virus, not chemical exposure.
Animal evidence for thyroid follicular cell tumors was slight based on increases seen in significant
pairwise increases of adenomas or carcinomas in female F344/N rats with a significant dose-response
trend but only marginal increases in male rats and no increase in B6C3F1 mice (NTP. 1991b). Although
U.S. EPA (1998a) notes that thyroid tumors in animal studies cannot be completely dismissed as a
hazard for humans, it appears that that rodents are more sensitive than humans to thyroid follicular cell
tumors induced by thyroid-pituitary disruption and thyroid stimulating hormone hyperstimulation
(Dybing and Sanner. 1999; U.S. EPA 1998a). There is also slight evidence in animals for harderian
gland adenoma or carcinoma based on increased incidence in female B6C3F1 mice at the highest dose
only, but no increased incidence in rats or male B6C3F1 mice (NTP. 1991b). Finally, slight evidence in
animals exists for hepatocellular tumors based on a dose-related trend in tumor incidence in only in one
sex of one species (male B6C3F1 mice) (NTP. 1991b).
The mechanistic evidence for carcinogenesis is slight. Available data indicates that TCEP has little of
any genotoxic potential, but data are limited to assess in vivo genotoxicity. Limited additional data
indicate that TCEP may influence cell signaling related to proliferation, apoptosis, and ion transport,
induce oxidative stress, alter cellular energetics in kidney tissues and cells and in other cell types.
U.S. EPA's PPRTV (U.S. EPA. 2009) also concluded that TCEP is likely to be carcinogenic to humans
based on information from oral animal bioassays that included clear evidence of renal tubule cell
adenomas in F344/N rats in NTP (1991b). renal tubule adenomas and carcinomas in ddY mice in
Takada et al. (1989) as well as the rarity of these tumors. The PPRTV also describes evidence for other
tumors identified in these two bioassays as suggestive or equivocal.
The 2009 European Union Risk Assessment Report (ECB. 2009) concluded that TCEP has
carcinogenicity potential and cites the EU classification category 3 and R40—limited evidence of
carcinogenic effect. In contrast, the International Agency for Research on Cancer (IARC) designated
TCEP as not classifiable as to its carcinogenicity to humans in 1990 and again in 1999 (IARC. 2019).
5.2.5 Dose-Response Assessment
According to U.S. EPA's 2021 Draft Systematic Review Protocol (U.S. EPA. 2021a). hazard endpoints
that receive evidence integration judgments of demonstrates and likely would generally be considered
for dose-response analysis. Endpoints with suggestive evidence can be considered on a case-by-case
basis. Studies that received high or medium overall quality determinations (or low-quality studies if no
other data are available) with adequate quantitative information and sufficient sensitivity can be
compared.
There were no hazard outcome categories for which evidence demonstrates that TCEP causes the effect
in humans. Therefore, hazard outcomes that received likely judgements are the most robust evidence
integration decisions. The health effect with the most robust and sensitive POD among these likely
outcomes was used for risk characterization for each exposure scenario to be protective of other adverse
effects as described in the sections below.
Data for the dose-response assessment were selected from oral toxicity studies in animals. No acceptable
toxicological data were available by the inhalation route, and no PBPK models are available to
extrapolate between animal and human doses or between routes of exposure using TCEP-specific
information.
Page 287 of 638
-------
The PODs estimated based on effects in animals were converted to HEDs or CSFs for the oral and
dermal routes and HECs or IURs for the inhalation route. For this conversion, EPA used guidance from
U.S. EPA (2011b) to allometrically scale oral data between animals and humans. Although the guidance
is specific for the oral route, EPA used the same HEDs and CSFs for the dermal route of exposure as the
oral route because the extrapolation from oral to dermal routes is done using the human oral doses,
which do not need to be scaled across species. EPA accounts for dermal absorption in the dermal
exposure estimates, which can then be directly compared to the dermal HEDs.
For the inhalation route, EPA extrapolated the daily oral HEDs and CSFs to HECs and IURs using
human body weight and breathing rate relevant to a continuous exposure of an individual at rest. Based
on existing data (Herr et al.. 1991). absorption via the oral route may be greater than 95 percent.
Therefore, EPA assumed that absorption for the oral routes is 100 percent; there is no information
regarding absorption via the inhalation route, and therefore, EPA assumed 100 percent absorption via
this route. Therefore, no adjustment specific to absorption is needed for the oral and inhalation routes.
For consistency, all HEDs and the CSF are expressed as daily doses and all HECs are based on daily,
continuous concentrations (24 hours per day) using a breathing rate for individuals at rest. Adjustments
to exposure durations, exposure frequencies, and breathing rates are made in the exposure estimates used
to calculate risks for individual exposure scenarios.
Appendix K.3 presents information on dose derivation, calculations for each of the PODs, and route-to-
route extrapolations. Considerations regarding the BMD modeling process as well as modeling results
for likely as well as suggestive TCEP outcomes are presented in the supplemental file Benchmark Dose
Modeling Results for TCEP (U.S. EPA. 2024c). A comparison of the PODs for likely and suggestive
health outcomes is presented visually in exposure response arrays within Appendix N, with calculations
for these PODs in an Excel spreadsheet in the supplemental file Raman Health Hazard Points of
Departure Comparison Tables (U.S. EPA. 2024k).
5.2.5.1 Selection of Studies and Endpoints for Non-cancer Toxicity
EPA considered the suite of oral animal toxicity studies and likely individual adverse health effects
outcomes when considering non-cancer PODs for estimating risks for acute and intermediate/chronic
exposure scenarios, as described in Sections 5.2.5.1.1 and 5.2.5.1.2, respectively. Epidemiological
studies were summarized for the weight of scientific evidence. EPA selected studies and relevant health
effects based on the following considerations:
• Overall quality determinations;
• Exposure duration;
• Dose range;
• Relevance (e.g., what species was the effect in, was the study directly assessing the effect, is the
endpoint the best marker for the tox outcome?);
• Uncertainties not captured by the overall quality determination;
• Endpoint/POD sensitivity;
• Total UF; and
• Uncertainty and sensitivity of BMR selection from BMD modeling.
The following sections provide comparisons of the above attributes for studies and hazard outcomes for
each of these exposure durations and details related to the studies considered for each exposure duration
scenario.
Page 288 of 638
-------
5.2.5.1.1 Non-cancer Points of Departure for Acute Exposure
To calculate risks for the acute exposure duration in the risk evaluation, EPA used a daily HED of 9.46
mg/kg (NOAEL of 40 mg/kg) from a prenatal/postnatal neurodevelopmental toxicity study (Moser et al..
2015) based on very slight to moderate tremors within five days of dosing at 125 mg/kg-day in 13 dams.
EPA gave this study a high overall quality determination, and a UF of 30 was used for the benchmark
MOE during risk characterization.
Mice exhibited signs of neurotoxicity in other acute or short-term high-quality studies. In the NTP
(1991b) 16-day study, mice exhibited ataxia and convulsive movements within three days at the two
highest doses with a daily HED of 16.6 mg/kg; data were only qualitatively described. Pregnant mice
administered 940 mg/kg-day TCEP via oral gavage were languid, prostrate, and exhibited jerking
movements during GDs 7 through 14 with an HED of 125 mg/kg-day (Hazleton Laboratories. 1983).
The HED from Moser et al. (2015) is more sensitive.
Tilson et al. (1990) found that in addition to convulsions, female Fischer 344 rats exhibited
histopathological changes in the hippocampus and memory impairment in the Morris water maze after a
single oral gavage administration of 275 mg/kg and an HED of 65.0 mg/kg. Although EPA gave Tilson
et al. (1990) a high overall quality determination, the authors tested only a single dose level, which did
not allow a full understanding of the dose-response for TCEP. The POD is associated with greater
uncertainty because only a LOAEL was identified and a UF of 300 would be required for a benchmark
MOE analysis.
The high-quality intraperitoneal injection study by Umezu et al. (1998) provides qualitative support for
neurotoxicity; mice exhibited increased ambulatory activity at 100 and 200 mg/kg and 'light'
convulsions at 200 mg/kg after single administration of these doses. EPA did not consider this study to
be a candidate for the POD based on the exposure route.
Table 5-47 presents a comparison of the attributes of studies and hazard endpoints considered for the
intermediate exposure scenario and Table 5-48 summarizes the study PODs and pertinent information,
including HEDs and HECs. The bolded row represents the study and POD values used to calculate risks
for acute scenarios in the risk evaluation.
Overall, the tremors observed in Moser et al. (2015) represent a sensitive endpoint that could occur in
humans. The clinical signs of neurotoxicity (e.g., convulsions) were consistently observed across
acute/short-term studies.
Page 289 of 638
-------
Table 5-47. Comparison among Studies with Sensitive Neurotoxicity Endpoints Considered for
Acute Exposure Scenarios
Moser et al. (2015)
NTP (1991b)
Tilson et al. (1990)
Hazleton
Laboratories
(1983)
Overall Data
Quality
Determination
High
High
High
High
Exposure
Duration
Within 5 days
Within 3 days
1 day
8 days
Dose Range
12, 40, 125 mg/kg-
day (high dose
changed to 90 mg/kg-
day at 5 days)
0, 44, 88, 175, 350,
700 mg/kg-day
275 mg/kg
940 mg/kg-day
Relevance
Assumed to be
relevant to humans;
clearly adverse
Assumed to be
relevant to humans
(similar effect as
chosen POD);
clearly adverse
Assumed to be
relevant to humans
(similar effect as
chosen POD);
clearly adverse
Assumed to be
relevant to humans
(similar effect as
chosen POD);
clearly adverse
Uncertainties Not
Captured
Elsewhere
Effects observed only
at the highest dose
BMD modeling not
possible; only
qualitative outcome
information
available
Precision of POD is
limited because no
NOAEL was
identified
Precision of POD is
limited because no
NOAEL was
identified
Sensitivity of
POD for
exposure
scenario
Sensitive endpoint
with an identified
NOAEL
Less sensitive
Most sensitive when
considering
comparison with 300
benchmark MOE
Least sensitive
Total UF
30
30
300
300
Page 290 of 638
-------
Table 5-48. E
»ose-Response Analysis of Selected Studies Considered for Acul
te Exposure Scenarios
Target
Organ/
System
Species
Duration
Study POD/
Type (mg/kg)fl
Effect
HEC
(mg/m3)
[ppm]
HED
(mg/kg)
UFs
Reference
Overall
Quality
Determination
Neurotoxicity
Long
Evans rats
(dams)
5 days
NOAEL = 40
Tremors
51.5
[4.41]
9.46
UFA=3
UFH=10
Total UF=30
Moser et al.
(2015)
High
Neurotoxicity
B6C3Fi
mice
16 days
NOAEL = 125
Convulsions,
ataxia within 3
days
90.4
[7.75]
16.6
UFA=3
UFH=10
Total UF=30
NTP (1991b)
High
Neurotoxicity
Fischer 344
rats
(females)
1 day
LOAEL = 275
Convulsions
brain lesions,
behavior
changes
354
[30.3]
65.0
UFA=3
UFH=10
UFl = 10
Total UF=300
Tilson et al.
(1990)
High
Neurotoxicity
CD-I mice
(dams)
GD 7-14
LOAEL = 940
Jerking
movements,
languidity,
prostration
680
[58.3]
125
UFA=3
UFH=10
UFl =10
Total UF=300
Hazleton
Laboratories
(1983)
High
" The PODs are duration adjusted to 7 days per week; therefore, any PODs from studies that dosed for 5 days per week were multiplied by 5/7.
Page 291 of 638
-------
5.2.5.1.2 Non-cancer Points of Departure for Intermediate and Chronic Exposures
Figure 5-17 presents exposure response arrays of the HEDs for the likely hazard outcomes from the
studies considered for the intermediate and chronic HEDs. The HEDs are presented within the hazard
outcomes of reproductive toxicity (with developmental as comparison), kidney toxicity, and
neurotoxicity and ordered from lowest to highest to view relative sensitivities more easily.
Karyomegaly; 2 yr; mouse (M); NTP 1991b
Hyperplasia; 2 yr; rat (M); NTP 1991b
Absolute and relative kidney wt; 66 wk; rat (M); NTP 1991b
Relative kidney wt; 16 wk; mouse (M); NTP 1991b
Karyomegaly; 2 yr; mouse (F); NTP 1991b
Absolute and relative kidney wt; 16 wk; rat (M); NTP 1991b
Absolute kidney wt; 16 wk; mouse (M); NTP 1991b
Relative kidney wt; 16 d; mouse (F); NTP 1991b
snerating tubules, other histopathological changes; 28 d; rat (M); Taniai et al. 2012
No. of seminiferous tubules; 35 d; mouse (M); Chen et al. 2015
Testiscular testosterone; 35 d; mouse (M); Chen et al. 2015
Absolute and relative testes wt; 16 wk; mouse (M); NTP 1991b
Sperm count; 16 wk; mouse (M); Matthews et al. 1990
Task 4: Fertility and pregnancy index in Fl; 14 wk; mouse (M,F); NTP 1991a
Testes wt; 35 d; mouse (M); Chen et al. 2015
Task 2: Fertility, litter 5 in FO; Up to 18 wk; mouse (M,F); NTP 1991a
k 2: Days to litter 2 and days to litter 3 in FO; Up to 18 wk; mouse (M,F); NTP 1991a
;an wt changes & histopathology; Sperm parameters; Pregnancy & fertility indices;
18 wk; mouse (M,F); NTP 1991a [1]
Brain lesions; 2 yr; rat (F); NTP 1991b
Hippocampal lesions; 60 d; rat (F); Yang et al. 2018
Brain (hippocampal) necrosis; 16 wk; rat(F); NTP 1991b; Matthews et al. 1990
Changes in path length, Morris water maze; 60 d; rat (F); Yang et al. 2018
Ataxia, convulsions; 16 d; mouse (NS); NTP 1991b
Brain lesions; 2 yr; mouse (M); NTP 1991b
Serum cholinesterase activity; 16 wk; rat(F); NTP 1991b; Matthews et al. 1990
Clinical observations; 60 d; rat (F); Yang et al. 2018
Prostration, jerking movements, languidity; 8 d; mouse (F); Hazleton Labs 1983
Task 2: Live male Fl pups/litter; Up to 18 wk; mouse (M); NTP 1991a
Task 4: Live F2 pups/litter; 14 wk; mouse (M,F); NTP 1991a
Task 2: FO mean litters/pair; Uve total Fl pups/litter; Uve female Fl pups/litter;
Figure 5-17. Exposure Response Array for Intermediate and Chronic Exposure Durations by
Likely Hazard Outcomes (and Developmental Toxicity)
Page 292 of 638
-------
EPA is using Chen et al. (2015a). the 35-day study in adolescent mice, to estimate non-cancer risks for
both the intermediate and chronic exposure scenarios. The study received a high overall quality
determination, and the sensitive effect is a decrease in the numbers of seminiferous tubules (by 22 and
41 percent at 100 and 300 mg/kg-day, respectively) that is accompanied by absolute disintegration of
tubules and decreased testosterone levels and testes weights at 300 mg/kg-day.
EPA conducted BMD modeling, and several continuous BMD models adequately fit the seminiferous
tubule numbers, resulting in similar BMDL5s. The exponential 2 model fit resulted in the lowest Akaike
information criterion (AIC) and a good fit upon visual inspection. U.S. EPA (2024c) presents additional
details, including the fits for all seven continuous models that were run and BMDL values for BMRs of
five percent RD and one SD.
For continuous data, EPA's BMD Technical Guidance recommends modeling the data using a BMR of
one standard deviation (SD) (U.S. EPA 2012b). but lower response rates should be used when effects
are severe (e.g., frank). Thus, EPA used a BMR of 5 percent based on biological severity and identified
a BMDL5 of 21 mg/kg-day. The BMDLs for 1 SD and 10 percent were 61 and 43 mg/kg-day,
respectively.
As stated in EPA's Guidelines for Reproductive Toxicity Risk Assessment (U.S. EPA. 1996). human
males are particularly susceptible to chemicals that reduce numbers or quality of sperm. Chen et al.
(2015a) did not directly evaluate sperm numbers or quality but due to potential for the endpoint to affect
fertility, the magnitude of effects observed in the study, and the potential for human males to be more
susceptible than rodents, EPA considers the significant effect on seminiferous tubules (which help
produce, maintain, and store sperm) to be severe and thus, warrants use of a BMR of 5 percent.
BMRs of 5 percent were also used for other severe or frank effects in the TCEP risk evaluation,
including decreased live pups per litter and brain necrosis. Furthermore, use of 5 percent for
degeneration of seminiferous tubules has support from other authors, even at levels down to 5 percent
degeneration of the tubules (Blessinger et al.. 2020). When evaluating male phthalate syndrome,
Blessinger et al. (2020) used a BMR of 5 percent for all endpoints associated with zero to moderate
impacts on fertility. These endpoints included germ cell degeneration or depletion in seminiferous
tubules ranging from 5 to 75 percent (Blessinger et al.. 2020; Lanning et al.. 2002).
EPA calculated a daily HED of 2.79 mg/kg-day for Chen et al. (2015a) that accounts for allometric
scaling between mice and humans and is compared with a benchmark MOE of 30. HEDs for other
reproductive effects ranged from 9.51 to 93.1 mg/kg-day. Many are within an order of magnitude of
Chen et al. (2015a). The HEDs of 93.1 mg/kg-day are based on LOAELs that are 33 times greater (NTP.
1991a) and are used with a benchmark MOE of 300 instead of 30.
As noted in Section 5.2.3.1.2, hazard outcomes identified by Chen et al. (2015a) are supported by effects
on sperm, reproductive organ weight changes, and testes hyperplasia (NTP. 1991a. b; Matthews et al..
1990). Other reproductive and developmental outcomes were observed, including decreases in fertility
and live pups per litter in the continuous breeding toxicity study (NTP. 1991a).
There are uncertainties associated with using Chen et al. (2015a) for the POD. Other than minimal to
mild hyperplasia, histopathological changes in the testes were not routinely identified in other studies
(NTP. 1991a. b). However, Chen et al. (2015a) was conducted more than 20 years after the NTP studies
and some methods differed from older studies (e.g., preparation of tissues). Also, differences may reflect
Page 293 of 638
-------
use of different species or mouse strains, and in such cases, U.S. EPA (1996) recommends using the
most sensitive species in the absence of information to suggest otherwise.
There are limitations of Chen et al. (2015a)'s study design and the BMD modeling analysis. Doses for
this feeding study may be imprecise because information on body weight and food consumption were
not reported. In addition, the sample size is small and as sample size decreases, uncertainty in the true
response rate increases. Finally, although EPA considered BMD modeling as appropriate for this data
set, in part because the lowest dose tested was a LOAEL, the BMR of 5 percent is lower than the
biologically and statistically adverse responses observed in the study (22.2 and 40.7%). Overall,
however, the significant and severe effect on seminiferous tubules is of concern for human males, a
sensitive subpopulation.
Comparison of Studies Usedfor the Intermediate Exposure Scenario: In addition to Chen et al. (2015a).
EPA considered sensitive effects from other studies ranging from a few days to 60 days for the
intermediate POD that would be associated with a 30-day exposure scenario. Table 5-49 presents a
comparison of the attributes of multiple studies and hazard endpoints considered for the intermediate
exposure scenario. Table 5-50 provides details of the studies, including PODs from the study or from
dose-response modeling, HECs, and HEDs. The bolded row represents the study and POD values used
to calculate risks for intermediate and chronic scenarios in the risk evaluation.
HEDs for both Moser et al. (2015) and Yang et al. (2018a) are based on neurotoxicity, which are
relevant hazard outcomes observed across multiple studies and are within an order of magnitude of the
sensitive HED (2.79 mg/kg-day) from Chen et al. (2015a). In addition, they are oral gavage studies and
thus, dose levels are expected to be more precise compared with Chen et al. (2015a). a dietary study.
However, exposure durations (5 and 60 days) for these studies introduce some uncertainty regarding
applicability to the target 30-day exposure scenario compared with Chen et al. (2015a). a 35-day study.
Even though the HED from Chen et al. (2015a) is based on using a BMR below the observed data, other
intermediate study and endpoint candidates also have limitations related to dose-response relationships.
Moser et al. (2015) observed effects only at the highest dose, and therefore, the HED is based on a
NOAEL, not a BMDL that considers the full dose-response curve. Similarly, the lowest HED (11.8
mg/kg-day) from Yang et al. (2018a) is based on a NOAEL; a similar HED from Yang et al. (2018a) (13
mg/kg-day, based on a BMDL20 of 55.0 mg/kg-day) also results in some uncertainty given typical
variability in the modeled neurobehavioral endpoint.
Taniai et al. (2012a). a 28-day study resulting in kidney proximal tubule regeneration, has a relevant
hazard outcome and an exposure duration closer to the intermediate scenario. However, even less is
known about the dose-response relationship because the study used only a single dose level resulting in
a LOAEL and a benchmark MOE of 300 rather than 30 used with Chen et al. (2015a).
Overall, using Chen et al. (2015a) for the intermediate exposure scenario in which adolescent male rats
were evaluated during a potentially sensitive lifestage results in a sensitive POD for a relevant endpoint
for the risk evaluation. EPA considers this POD to be protective of other adverse effects identified in
TCEP toxicity studies.
Page 294 of 638
-------
Table 5-49. Comparison among Studies with Sensitive Endpoints Considered for Intermediate
Exposure Scenarios
Neurotoxicity
(Moser et al., 2015)
Neurotoxicity
(Yane et al., 2018a)
Reproductive
Toxicity
(Chen et al., 2015a)
Kidney Toxicity
(Taniai et al.,
2012a)
Overall Data
Quality
Determination
High
High
High
Medium
Exposure Duration
Within 5 days; less
applicable to
intermediate exposure
60 days; less
applicable to
intermediate exposure
35 days
28 days
Dose Range
12, 40, 125 mg/kg-day
(high dose changed to
90 mg/kg-day at 5
days)
50, 100, 250 mg/kg-
day
100, 300 mg/kg-day
350 mg/kg-day
Relevance
Endpoint assumed to
be relevant to humans
Endpoint assumed to
be relevant to humans
Endpoint assumed to
be relevant to human
male reproduction
(U.S. EPA. 1996)
Endpoint assumed to
be relevant to
humans
Uncertainties Not
Captured
Elsewhere
Dose-response less
precise: Use of
NOAEL
Dose-response less
precise: Use of
NOAEL);
Neurobehavioral
outcomes (BMR of
20%) had a similar
HED (13 mg/kg-day)
but effect is typically
variable
Dose precision
unclear: dietary study
and no information
on food consumption
or body weight
Lack of
understanding of
dose response and
greater uncertainty
due to use of single
dose level resulting
in a LOAEL
Sensitivity of
Endpoint and POD
Within an order of
magnitude of the most
sensitive endpoint
Within an order of
magnitude of the most
sensitive endpoint
Most sensitive
endpoint for the
intermediate scenario
Less sensitive
endpoint but is used
with a larger
benchmark MOE
Total UF/
Benchmark MOE
30
30
30
300
Uncertainty/
Sensitivity of
BMR Selection
N/A
N/A
BMR of 5% is lower
than responses in
study
N/A
Page 295 of 638
-------
Table 5-50. Dose-Response Analysis of Selected Studies Considered for Intermediate Exposure Scenarios
Target Organ/
System
Species
Duration
Study POD/
Type
(mg/kg-day)
Effect
HEC
(mg/m3)
[ppm]
HED
(mg/kg-day)
UFs
Reference
Overall
Quality
Determination
Reproductive
Toxicity
ICR mice
(males)
35 days
BMDLs = 21"
Decreased
numbers of
seminiferous
tubules
14.9
[1.27]
2.73
UFA=3
UFH=10
Total UF=30
Chen et al.
(2015a)
High
Neurotoxicity
Sprague-
Dawley
rats
(females)
60 days
NOAEL = 50
Hippocampal
lesions
64.3
[5.51]
11.8
UFA=3
UFH=10
Total UF=30
Yang et al.
(2018a)
High
Kidney
Toxicity
F344 rats
(males)
28 days
LOAEL = 350
Regenerating
tubules in
kidneys
450
[38.6]
82.8
UFA=3
UFH=10
UFL=10
Total UF=300
Taniai et al.
(2012a)
Medium
11 The BMDL based on 1SD is 61.2 mg/kg-day.
Page 296 of 638
-------
Comparison of Studies and Hazard Outcomes for the Chronic Exposure Scenario: EPA generally
considers chronic studies to be those with exposure durations of > 10 percent of a lifetime. For TCEP,
these studies include the 16- and 18-week and 2-year NTP studies in rats and mice (NTP. 1991b). Also,
many of the endpoints in the RACB study (NTP. 1991a) (especially the crossbreeding and second-
generation effects) were measured after chronic exposure. Table 5-51 presents a comparison of the
attributes of sensitive endpoints from studies considered for the chronic exposure scenario, and Table
5-52 provides study details including PODs from the study or BMD modeling results, HECs, and HEDs.
Although it is a study with a shorter exposure duration, EPA chose Chen et al. (2015a) for the chronic
exposure scenarios because it resulted in an HED that is more sensitive (2.79 mg/kg-day) than most
longer-term results and covers a potentially sensitive lifestage (adolescence).
Use of the shorter duration study by Chen et al. (2015a). however, does lend uncertainty to the risk
evaluation because other longer-term studies are not as sensitive and because it is uncertain whether the
POD would be lower if Chen et al. (2015a) extended the exposure duration.
For the endpoints that resulted in likely evidence integration conclusions, most chronic studies received
high overall quality determinations. There were a few exceptions. EPA gave medium overall quality
determinations to the sperm morphology and vaginal cytology results reported in the 16- and 18-week
NTP studies (Matthews et al.. 1990). primarily based on limited information regarding methods and
results. Clinical observations described by NTP (1991b) for the 16- and 18-week studies in mice and rats
received uninformative overall quality determinations due to the lack of reasonably available
quantitative information for these effects.
The single chronic endpoint more sensitive than Chen et al. (2015a) was increased relative kidney
weights for female rats from the 16-week NTP study, with an HED of 1.75 mg/kg-day (NTP. 1991b).
However, EPA considered the changes in kidney weights for TCEP less relevant for predicting kidney
toxicity than other endpoints (i.e., kidney histopathology) because they were not consistently observed;
female rats had increased relative kidney weights after 16 weeks but not after 66 weeks, and female
mice had increased weights at 16 days but not at 16 weeks or the 66-week sacrifice. In addition, kidney
weight changes did not correspond to histopathology changes (NTP. 1991b).
Histopathology is a more reliable endpoint for kidney effects and was observed in the 2-year studies
(NTP. 1991b); daily HEDs associated with hyperplasia and karyomegaly ranged from 5.49 to 14.2
mg/kg-day; most are within a factor of three of Chen et al. (2015a) and 14.2 mg/kg-day is roughly five
times higher.
Neurotoxicity was consistently observed across chronic studies with HEDs ranging from 7.43 to 22.8
mg/kg-day. These HEDs are all within an order of magnitude of Chen et al. (2015a).
The comparison of HEDs with reproductive endpoints described earlier and the comparisons with
kidney and neurotoxicity endpoints observed in the chronic studies demonstrates some consistency
across endpoints with respect to potency. These co-critical endpoints lend strength to using the sensitive
endpoint from Chen et al. (2015a) for the chronic duration.
Similar to Chen et al. (2015a). only two dose groups (44 and 88 mg/kg-day) were used in the NTP
(1991b) 2-year studies associated with the most sensitive of the kidney and neurotoxic effects, which
somewhat limits the understanding of the dose response relationship for these endpoints.
Page 297 of 638
-------
Overall, the HED from Chen et al. (2015a) associated with a relevant hazard outcome is protective of
other observed adverse effects from chronic exposure to TCEP that include neurotoxicity and kidney
histopathological effects.
Table 5-51. Comparison among Studies with Sensitive Endpoints Considered for Chronic
Exposure Scenarios
Neurotoxicity
(NTP, 1991b)
Reproductive Toxicity
(Chen et al., 2015a)
Kidney
(NTP. 1991b)
Overall Data Quality
Determination
High
High
High
Exposure Duration
2-year; chronic
3 5-day; intermediate (<
chronic)
2-year; chronic
Dose Range
44, 88 mg/kg-day
100, 300 mg/kg-day
44, 88 mg/kg-day
Relevance
Endpoint assumed to be
relevant to humans
Endpoint assumed relevance
to human male reproduction
(U.S. EPA. 1996); severity
identified
Endpoint assumed to be
relevant to humans
Uncertainties Not
Captured Elsewhere
Dose-response less
precise (use of NOAEL)
Dose precision unclear based
on dietary study with no
information on food
consumption or body weight
changes
Some inconsistencies
between kidney weight
changes and histopathology
Sensitivity of Endpoint
and POD
Most sensitive among
chronic neurotoxic
effects
Most sensitive across hazard
outcomes (except increased
kidney weight in 16-week
study)
Most sensitive among
chronic histopathological
kidney effects; 16-week
kidney weight change more
sensitive
Total UF
30
30
30
Uncertainty/Sensitivity
of BMR Selection
N/A
BMR of 5 percent, predicted
BMD and BMDL values are
lower than doses associated
with responses observed in the
study
BMR of 10 percent
Page 298 of 638
-------
Table 5-52. E
•ose-Response Analysis of Selected Stut
ies Consideret
for Chronic Exposure Scenarios
Target Organ
System
Species/Sex
Exposed
Duration
Study
POD/Type
(mg/kg-day)
Effect
HEC
(mg/m3)
[ppm]
HED
(mg/
kg-day)
UFs
Reference
Overall
Quality
Determination
Reproductive
Toxicity
ICR mice
(male)
35 days
BMDLs = 21"
Decreased
numbers of
seminiferous
tubules
14.9
[1.27]
2.73
UFA=3
UFH=10
Total UF=30
Johnson et al.
(2003)
Chen et al.
(2015a)
High
Neurotoxicity
F344 rats
(female)
2 years
NOAEL = 31.4
Brain lesions
40.4
[3.46]
7.43
UFA=3
UFH=10
Total UF=30
NTP (1991b)
High
Kidney
Toxicity
F344 rats
(female)
2 years
BMDLi o=23.2
Renal tubule
hyperplasia
30
[2.6]
5.49
UFA=3
UFH=10
Total UF=30
NTP (1991b)
High
" The BMDL based on 1SD is 61.2 mg/kg-day.
Page 299 of 638
-------
5.2.5.1.3 Uncertainty Factors Used for Non-cancer Endpoints
For the non-cancer health effects, EPA used a total uncertainty factor (UF) of 30 for the benchmark
MOEs for acute, intermediate, and chronic exposure durations for all exposure routes among studies that
are used to estimate risks. Other endpoints that used LOAELs for which EPA used a LOEAL-to-
NOAEL UF of 10 and a total benchmark MOE of 300.
1. Interspecies Uncertainty Factor (UFa) of 3
EPA uses data from oral toxicity studies in animals to derive relevant HEDs, and U.S. EPA
(2011a) recommends allometric scaling (using the 3/4 power of body weight) to account for
interspecies toxicokinetics differences for oral data. When applying allometric scaling, EPA
guidance recommends reducing the UFa from 10 to 3. The remaining uncertainty is associated
with interspecies differences in toxicodynamics. EPA also uses a UFa of 3 for the inhalation
HEC and dermal HED values because these values are derived from the oral FLED.
2. Intraspecies Uncertainty Factor (UFh) of 10
EPA uses a default UFh of 10 to account for variation in sensitivity within human populations
due to limited reasonably available information regarding the degree to which human variability
may impact the disposition of or response to, TCEP.
3. LOAEL-to-NOAEL Uncertainty Factor (UFl) of 1 or 10
The PODs chosen to calculate risks were either NOAELs or BMDL values and therefore, EPA
used a UFl of 1. EPA compared these values with other endpoints based on LOAELs, which
used a UFl of 10 to account for the uncertainty inherent in extrapolating from the LOAEL to the
NOAEL.
U.S. EPA (1993a) and U.S. EPA (2002b) further discuss use of UFs in human health hazard dose-
response assessment.
5.2.5.2 Selection of Studies and Endpoint Derivation for Carcinogenic Dose-Response
Assessment
EPA considered the kidney tumors for derivation of toxicity values for the risk calculations based on the
evidence integration conclusion that the tumors are sensitive and robust, and that cancer is likely to be
caused by TCEP. The selection of representative cancer studies and tumors for dose-response analysis is
described below based on the following considerations:
• Overall quality determination;
• Sufficiency of dose-response information;
• Strength of the evidence supporting the associated tumor type;
• MO A conclusions;
• Relevance (e.g., what species was the effect in, was the study directly assessing the effect, is the
endpoint the best marker for the tox outcome?);
• Uncertainties not captured by the overall quality determination; and
• Endpoint sensitivity.
Rodent bioassays identify increased incidences of kidney tumors in male F344/N rats, with a lower
increase in female rats (NTP. 1991b). Treatment-related kidney tumors were also observed after two
years in male B6C3Fi mice (NTP. 1991b). EPA gave NTP (1991b) a high overall quality determination.
Based on a lack of adequate information on mechanisms or temporality and lack of dose-response data
for precursor lesions to consider an alternate dose-response using a threshold analysis, EPA used linear
Page 300 of 638
-------
low-dose extrapolation to estimate risks. U.S. EPA's PPRTV also used linear low-dose extrapolation in
the absence of specific mechanistic information.
EPA used the multistage models available in the BMD software and adjusted the data for mortality by
using animals still alive on the first day of cancer incidence. Therefore, animals dying from other causes
were not included in the analysis. For both male and female rats, kidney tumor incidence data
adequately fit one or both multistage models and tumors in males (adenomas and carcinomas) resulted
in the more sensitive CSF (0.0058 per mg/kg-day). The IUR is based on daily, continuous
concentrations (24 hours per day) using a breathing rate for individuals at rest. Adjustments to exposure
durations, exposure frequencies, and breathing rates are made in the exposure estimates used to calculate
risks for individual exposure scenarios.
Table 5-53 presents the cancer PODs for modeled renal tumors. Because EPA has not concluded that
TCEP acts via a mutagenic mode of action, an age-dependent adjustment factor (ADAF) (U.S. EPA
2005c) was not applied when estimating cancer risk for kidney tumors from TCEP exposure. EPA did
not use CSFs for combined tumors (across multiple target organs) for the risk evaluation but focused on
the tumors with the most robust evidence from the animal data.
See Appendix K.3 for dose-response derivation, including details on route-to-route extrapolation.
Considerations regarding the BMD modeling process for cancer and results are presented in Benchmark
Dose Modeling Results for TCEP (U.S. EPA. 2024c).
EPA did not use CSFs for combined tumors (across multiple target organs) for the risk evaluation but
focused on the tumors with the most robust evidence from the animal data.
Table 5-53. Dose-Response Analysis of Kidney Tumors" for Lifetime Exposure Scenarios
Tumors
Species (Sex)
Oral/Dermal
CSFfl6
IURfl
Extra Cancer Risk
Benchmark
Renal tubule
adenomas or
carcinomas
F344 rats (male)
0.0245 per mg/kg-
day
0.00451 permg/m3
(0.0526 per ppm)
1E-04 (occupational)
1E-04 to 1E-06 (consumer,
general population)
Renal tubule
adenomas
F344 rats (female)
0.0220 per mg/kg-
day
0.00404 permg/m3
(0.0472 per ppm)
° CSFs and IURs were derived based on continuous exposure scenarios; CSFs from BMD modeling prior to allometric
scaling were 0.0058 and 0.0052 per mg/kg-day for male and female rats, respectively.
b U.S. EPA's PPRTV (U.S. EPA. 2009) calculated an oral CSF of 0.02 oer ma/ka-dav. also based on increased renal
tubule adenomas or carcinomas in male rats from NTP (1991b).
5.2.6 Weight of Scientific Evidence Conclusions for Human Health Hazard
EPA used information described in previous sections and additional factors to arrive at confidence levels
for key human health hazard outcome and exposure duration combinations. Evidence integration was
conducted by considering Bradford Hill criteria, as described in Section 5.2.1, with conclusions for
TCEP described in Sections 5.2.3 and 5.2.4.4. Only likely evidence integration conclusions were
considered for dose-response and given weight of evidence confidence levels. As outlined in Section
5.2.5, factors in addition to the Bradford Hill criteria were considered when choosing studies for dose-
response modeling and for each exposure scenario (acute, intermediate, and chronic).
Page 301 of 638
-------
The current section combines the evidence integration conclusions, factors already considered during
dose-response as well as additional considerations (see Section 5.2.6.1) to choose the overall hazard
confidence levels:
• Evidence integration conclusion (from Appendix L)
o Demonstrates is rated as +++
o Likely is rated as ++
o Suggests is rated as +
• Selection of key endpoint and study
• Relevance to exposure scenario
• Dose-response considerations
• PESS sensitivity
Section 5.2.6.2 presents a summary table of confidence for each hazard endpoint and exposure duration.
5.2.6.1 Strengths, Limitations, Assumptions, and Key Sources of Uncertainty for the
Hazard Identification and Selection of PODs for Human Health Hazard
Assessment
5.2.6.1.1 Acute Non-cancer
Evidence Integration Conclusions
Clinical signs of neurotoxicity, histopathological changes in the brain, and neurobehavioral changes
measured in multiple studies were considered for the acute exposure scenario. EPA concluded that
TCEP likely causes neurotoxicity in humans under relevant exposure circumstances and assigned high
overall quality determinations to all acute studies considered.
Selection of the Key Endpoint and Study
Several human epidemiological studies evaluated neurotoxicity. There appears to be possible support for
TCEP's association of prenatal exposure and impairment in cognitive abilities based on effects on IQ in
lower SES children (Percy et al.. 2022). The tremors observed in Moser et al. (2015) and similar
neurotoxic effects in other studies are key because they are adverse, and neurotoxicity is consistently
observed among acute and longer-term studies.
Offspring do not appear to be more sensitive for developmental neurotoxicity up to 90 mg/kg-day39 after
exposure of pregnant rats during gestation and the early postnatal period based on results from Moser et
al. (2015). Viability and growth of offspring were also not affected after pregnant mice were dosed with
940 mg/kg-day (Hazleton Laboratories. 1983).40
Relevance to Exposure Scenario
The candidate studies and endpoints for acute exposure identified neurotoxicity after one to eight days,
and EPA considered these durations relevant for the acute exposure scenario. The study, Moser et al.
(2015). chose to calculate risks and identified tremors within five days of exposure. There is some
uncertainty for this human exposure scenario given the lack of TCEP-specific information or models
(e.g., PBPK models) to extrapolate from animals to humans. EPA also extrapolated from oral HEDs to
inhalation HECs and dermal HEDs, which lends uncertainty for these routes. It is not known whether
39 The study began with a dose of 125 mg/kg-day, which was lower to 90 mg/kg-day after 5 days due to toxicity in dams at
the highest dose.
411A prenatal study in Wistar rats (Kawashima et al.. 1983) in a foreign language will be translated it into English and
evaluated for the risk evaluation.
Page 302 of 638
-------
these assumptions for the chosen POD would lead to over- or underprediction of risk from acute
exposure.
Dose-Response Considerations
None of the studies considered for acute exposure could be modeled using BMD models due to limited
reasonably available dose-response information. EPA identified a NOAEL from Moser et al. (2015). and
effects were seen only at the highest dose. The other acute studies also identified only a NOAEL or
LOAEL with effects observed only at the highest dose or the only dose in the study.
Susceptible Subpopulations
Moser et al. (2015) evaluated effects in pregnant female rats. Given the lower HED for this study
compared with other acute studies, pregnant dams may be a susceptible subpopulation. However,
uncertainties exist because of limited dose response information for other studies. Non-pregnant female
rats are also shown to be a sensitive species and sex for neurotoxicity in longer-term studies as identified
in NTP (1991b). Offspring, as noted earlier, were not identified as more sensitive to neurotoxicity or
other effects from gestational and postnatal exposure of the dams.
5.2.6.1.2 Intermediate and Chronic Non-cancer
Evidence Integration Conclusions
EPA considered multiple animal toxicity studies and multiple hazard outcomes - reproductive toxicity,
neurotoxicity, and kidney toxicity - for the intermediate and chronic exposure scenarios. Of the animal
toxicity studies/endpoints, EPA assigned high quality determinations except Taniai et al. (2012a). which
EPA gave a medium quality determination.
Selection of the Key Endpoint and Study
The nature of the effect chosen for calculating risks—differences in numbers and degeneration of
seminiferous tubules identified by Chen et al. (2015a)—is considered adverse, and the fertility of human
males is known to be sensitive to changes in sperm numbers and quality (U.S. EPA. 1996).
Neurotoxicity and kidney toxicity were also observed consistently among studies and HEDs were often
within an order of magnitude of each other.
The effects of Chen et al. (2015a) were the most sensitive after intermediate exposure. Increased relative
kidney weight was most sensitive after chronic exposure, but EPA considered these weight changes less
predictive of kidney toxicity due to inconsistencies between intermediate and longer-term studies and
lack of correlation with histopathology and clinical chemistry results in many cases.
Using Chen et al. (2015a) does lead to uncertainty because other studies did not report decreased
numbers or disintegration of seminiferous tubules; furthermore, related male reproductive effects were
only seen at higher doses in other studies. However, male reproduction was consistently affected in
several studies along with fertility and offspring viability. Thus, EPA considers the sensitive effects in
Chen et al. (2015a) to be relevant and differences might be due to species, test methods, or lifestage.
There are several considerations that lend uncertainty as to whether risks could be underpredicted using
this POD. These include lack of human data specific for male reproductive effects; the known sensitivity
of human males to reproductive insults; and uncertainty about certain sensitive effects that could not be
considered for a POD due to an error in the results presented in the continuous breeding study (NTP.
1991a) or lack of full reports (see Section 5.2.3.1.2).
Page 303 of 638
-------
There is some uncertainty as to whether this POD is protective of a full range of effects. For example,
chronic studies did not evaluate neurobehavioral batteries. In addition, EPA did not locate any studies
that investigated TCEP's association with acoustic startle responses or social behaviors.
Relevance to Exposure Scenarios
The 35-day exposure used by Chen et al. (2015a) is more relevant than the shorter and longer studies of
5 or 60 days (e.g., Moser et al. (2015) and Yang et al. (2018a)) for the intermediate exposure scenario,
which EPA defines as a 30-day exposure for this risk evaluation. Although the 28-day Taniai et al.
(2012a) study is well-suited for intermediate exposures, other study aspects limit its suitability,
including testing at only 350 mg/kg-day.
There is inherent uncertainty in assuming that a 35-day toxicity study in rodents during male
adolescence is applicable to a similar exposure duration in human adolescent males for the endpoint of
decreased numbers of seminiferous tubules.
Using Chen et al. (2015a) to represent chronic exposure durations adds uncertainty to the risk
evaluation. If the specific effect identified by Chen et al. (2015a) were measured in a chronic study in
the same species starting in adolescence, the POD could be more sensitive. Therefore, it is possible that
risks might be under-predicted. Yet, the available chronic HEDs were less sensitive than Chen et al.
(2015a).
For all studies and endpoints, no TCEP-specific information was available for extrapolation to humans
and EPA relied on allometric scaling based on BW3 4. Route-to-route extrapolation to inhalation HECs
and dermal HEDs results in additional uncertainty. EPA cannot predict whether the assumptions
regarding route extrapolation for the chosen POD would lead to over- or underprediction of risk from
intermediate exposure for the dermal route.41
Dose-Response Considerations
Chen et al. (2015a) fed TCEP to rats in a dietary study and did not report information on food
consumption. Thus, EPA does not know the precise doses received by the rats. However, the data
adequately fit several BMD models based on statistics and visual inspection and resulted in similar
BMDLs among the fit models. Also, use of the BMDL allowed EPA to use a relatively low total UF of
30. Given the severity of the effect (large percent decrease in numbers of tubules and significant
degeneration), EPA chose a BMR of 5 percent.
Although other intermediate studies with relevant sensitive effects used three treatment levels (vs. two
for Chen et al. (2015a)). EPA identified limitations for these other studies that included the inability to
conduct BMD modeling, use of only one dose (with LOAEL only) or an effect seen only at the highest
dose. Sensitive chronic neurotoxic and kidney effects are from studies with two treatment levels;
neurotoxicity could not be modeled (and only aNOAEL is available) but kidney hyperplasia could be
modeled and yielded an appropriate BMDL.
41 Data from Shepel'skaia and Dvshginevich (1981) (as cited in (NTP. 1991a)) suggests that testes effects occurred by
inhalation. If they are adverse and occurred at either low or high dose (0.5 or 1.5 mg/m3), the effect is more sensitive than the
extrapolated HEC value from Chenet al. (2015a). Shepel'skaia and Dvshginevich (1981) was not readily available to EPA
and appears to be only an abstract. Thus, EPA cannot consider Shepel'skaia and Dvshginevich (1981) for use in this risk
evaluation.
Page 304 of 638
-------
Susceptible Subpopulations
Chen et al. (2015a) evaluated a sensitive sex lifestage (male adolescent mice) and identified a sensitive
POD among critical endpoints. Other studies and endpoints considered for intermediate and chronic
exposure identified sexes that might be more sensitive to certain effects. For example, female rats were
more sensitive for neurotoxicity.
5.2.6.1.3 Cancer
Evidence Integration Conclusions
EPA concludes that TCEP is likely to be carcinogenic to humans using guidance from U.S. EPA's
Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005b) based on information from a high-
quality study (NTP. 1991b).
Selection of the Key Endpoint and Study
Of the organs that exhibited tumors in NTP (1991b). EPA used the tumor type with the most robust
evidence - kidney adenomas and carcinomas - and used a CSF that was the most sensitive among
modeled kidney tumor incidence.
EPA considers increased incidence of renal tubule adenomas and carcinomas to be adverse, relevant to
humans, and representative of a continuum of benign to malignant tumors and was the only target organ
with robust evidence of increased tumors. There is possible support for TCEP's association with thyroid
tumors in humans based on a case control study (Hoffman et al.. 2017). but the evidence is mixed based
on a lack of association in a second study (Liu et al.. 2022).
Of the kidney tumors, NTP (1991b) identified primarily adenomas and only one carcinoma. Thus, the
risk of malignant tumors is less certain; if humans are like rodents, use of the CSF from NTP (1991b)
could result in an over prediction of malignant cancer. However, if humans are more sensitive and
develop malignancies sooner, risks may be underpredicted.
Relevance to Exposure Scenarios
NTP (1991b) is a 2-year bioassay and is relevant for chronic exposures in humans. However, like non-
cancer endpoints, use of allometric scaling among species and route-to-route extrapolation to inhalation
HECs and dermal HEDs leads to some uncertainties and the impacts on risks are unknown.
Dose-Response Considerations
There is no complete understanding regarding mechanism(s) of cancer and there is also a lack of
appropriate precursors to cancer in the available in vivo studies with respect to temporality and dose
response (e.g., the single dose used by Taniai et al. (2012a) is higher than doses associated with tumors).
Therefore, EPA used linear low dose extrapolation a BMDLio. Because direct mutagenicity is not likely
to be the predominant MO A, using linear low dose extrapolation may be a health conservative analysis.
Use of tumor data for only one target organ (i.e., not combining incidence with other target organ
tumors) may result in some underestimation of risk, however. Therefore, the net effect of the dose-
response modeling, considering the benchmark risk levels used in the risk evaluation (1 in 10,000 to 1 in
1,000,000) is not known.
Page 305 of 638
-------
Susceptible Subpopulations
The single human study identified regarding TCEP exposure and thyroid cancer did not identify a
specific susceptible subpopulation (Hoffman et al.. 2017). Availability of a high-quality animal study
using two species and both sexes suggests possible sensitivities by sex (e.g., higher incidence of kidney
tumors in male rats).
The dose-response model applied to animal tumor data employed low-dose linear extrapolation, and this
assumes any TCEP exposure is associated with some positive risk of getting cancer. However, EPA did
not identify specific human groups that are expected to be more susceptible to cancer following TCEP
exposure even though there is likely to be variability in susceptibility across the human population.
Other than relying on animal tumor data for the more sensitive sex, the available evidence does not
allow EPA to evaluate or quantify the potential for increased cancer risk in specific subpopulations.
Given that a mutagenic mode of action is unlikely, EPA does not anticipate greater cancer risks from
early life exposure to TCEP.
5.2.6.2 Human Health Hazard Confidence Summary
Table 5-54 summarizes the confidence ratings for each factor for key human health hazards considered
for acute, intermediate, chronic, and lifetime exposure scenarios. The bolded rows are the health
endpoints for each exposure scenario used to calculate risks. Alternate PODs for health outcomes are not
bolded in the table.
Table 5-54. Confidence Summary for Human Health Hazard Assessment
Hazard
Domain
Evidence
Integration
Conclusion
Selection of Most
Critical Endpoint
and Study
Relevance to
Exposure
Scenario
Dose-Response
Considerations
PESS
Sensitivity
Overall
Hazard
Confidence
Acute non-cancer
Neurotoxicity
+ +
+ + +
+ +
+ +
+ +
Moderate
Intermediate non-cancer
Reproductive
+ +
+ +
+ + +
+
+ +
Moderate
Neurotoxicity
+ +
+
+ +
+ +
+ +
Moderate
Kidney
+ +
+
+ + +
+
+
Moderate
Chronic non-cancer
Reproductive
+ +
+ +
+
+
+ +
Moderate
Neurotoxicity
+ +
+
+ + +
+ +
+ +
Moderate
Kidney
+ +
+
+ + +
+ +
+
Moderate
Cancer
Kidney Cancer
+ +
+ +
+ + +
+ +
+ +
Moderate
+ + + Robust confidence suggests thorough understanding of the scientific evidence and uncertainties. The supporting
weight of scientific evidence outweighs the uncertainties to the point where it is unlikely that the uncertainties could have a
significant effect on the hazard estimate.
+ + Moderate confidence suggests some understanding of the scientific evidence and uncertainties. The supporting
scientific evidence weighed against the uncertainties is reasonably adequate to characterize hazard estimates.
+ Slight confidence is assigned when the weight of scientific evidence may not be adequate to characterize the scenario, and
when the assessor is making the best scientific assessment possible in the absence of complete information. There are
additional uncertainties that may need to be considered.
5.2.7 Toxicity Values Used to Estimate Risks from TCEP Exposure
After considering hazard identification and evidence integration, dose-response evaluation, and weight
of scientific evidence of POD candidates, EPA chose two non-cancer endpoints for the risk evaluation—
Page 306 of 638
-------
one for acute exposure scenarios and a second one for intermediate and chronic scenarios (Table 5-55).
Cancer risks were estimated using increased kidney tumors in male rats (Table 5-56). HECs and IURs
are based on daily continuous (24-hour) exposure and HEDs and CSFs are daily values. All studies
received high overall quality determinations.
Page 307 of 638
-------
Table 5-55.
Von-cancer HECs and HE
)s Used to Estimate Risks
Exposure
Scenario
Target Organ
System
Species
(Sex)
Duration
POD
(mg/kg-day)
Effect
HEC
(mg/m3)
[ppm]
HED
(mg/
kg-day)
Benchmark
MOE
Reference(s)
Acute
Neurotoxicity
Long Evans
rats (dams)
5 days
NOAEL =
40
Tremors
51.5
[4.41]
9.46
UFA = 3
UFH = 10
Total UF = 30
Moser et al.
(2015)
Intermediate
and Chronic
Reproductive
Toxicity
ICR mice
(male)
35 days
bmdl5 =
21
Decreased
seminiferous
tubules
14.9
[1.27]
2.73
UFA=3
UFH=10
Total UF = 30
Johnson et al.
(2003)
Chen et al.
(2015a)
Table 5-56.
Cancer IUR and CSF Uset
to Estimate Risks
Exposure
Scenario
Target Organ
System
Species
(Sex)
Duration
POD
(mg/kg-day)
Effect
IUR
(per mg/m3)
[per ppm]
CSF
(per mg/
kg-day)
Benchmark
Risk Levels
Reference
Chronic/
Lifetime
Kidney tumors
Fischer
344/N rats
(male)
2 years
CSF from
BMD model
= 0.0058 per
mg/kg-day
Increased
renal tubule
adenomas or
carcinomas
0.00451
[0.0526]
0.0245
1E-04
(occupational)
1E-04 to 1E-06
(consumer,
general
population)
NTP (1991b)
Page 308 of 638
-------
Table 5-57. General Population Acute Drinking Water (Oral Ingestion) Non-cancer Risk Summary
cou
OES
Acute Oral Non-cancer MOEs
UFs = 30
Drinking Water
Drinking Water (Diluted)
Life Cycle
Stage/Category
Subcategory
Adult
(>21 yr)
Infant
(21 yr)
Infant
(
-------
Table 5-58. Acute Fish Ingestion Non-cancer Risk Summary Based on 50th Percentile Flow of Harmonic Mean
cou
OES
Acute Oral Non-cancer MOEs
UFs = 30
General Population
Subsistence Fishers
Tribes
(Current IR)a
Tribes
(Heritage IR)A
Life Cycle Stage/
Category
Subcategory
BAF
2,198
BAF 109
BAF 2,198
BAF 109
BAF
2,198
BAF 109
BAF
2,198
BAF 109
Manufacturing/
Import
Import
Repackaging
18
363
3
57
2
37
<1
5
Processing/
Incorporation into
formulation, mixture,
or reaction product
Paint and coating
manufacturing
Incorporation into paints
and coatings - 1 -part
coatings
4
82
1
13
<1
8
<1
1
Incorporation into paints
and coatings - 2-part
reactive coatings
4
90
1
14
<1
9
<1
1
Polymers used in
aerospace
equipment and
products
Formulation of TCEP-
containing reactive resin
3
65
<1
10
<1
7
<1
1
Commercial
use/Laboratory
chemicals
Laboratory
chemical
Use of laboratory
chemicals
450
9,067
70
1,411
46
930
6
122
Commercial and
Industrial use/Paints
and coatings
Paints and
coatings
Use of paints and
coatings at job sites
8
154
1
24
1
16
<1
2
" Current fish consumption rate at 216 g/day based on survey of Suquamish Indian Tribe in Washington (see Section 5.1.3.4.4)
4 Heritage fish consumption rate at 1,646 g/day based on study of Kootenai Tribe in Idaho (see Section 5.1.3.4.4)
Page 310 of 638
-------
Table 5-59. General Populai
ion Chronic Water and Soil Ingestion Non-cancer Risk Summary
cou
OES
Chronic Non-cancer Oral MOEs
UFs = 30
Life Cycle
Stage/C atego ry
Subcategory
Drinking
Water
(Diluted)
Drinking
Water
Drinking
Water (via
Leaching to
Groundwater)
Ambient
Water
(Incidental
Ingestion)
Soil Intake
(50th)
at 100 m
Soil Intake
(95th)
at 100 m
Soil Intake
(50th)
at 1,000 m
Soil Intake
(95th)
at 1,000 m
Manufacturing/
Import
Import
Repackaging
1.64E08
1.05E05
N/A
2.11E05
2.20E10
5.15E09
1.73E12
4.03E11
Processing/
Incorporation
into formulation,
mixture, or
reaction product
Paint and coating
manufacturing
Incorporation into
paints and
coatings - 1 -part
coatings
4.40E07
23,728
2.12E06
4.89E04
7.02E08
1.64E08
7.95E10
1.86E10
Incorporation into
paints and
coatings - 2-part
reactive coatings
4.85E07
26,171
N/A
5.39E04
4.85E09
1.13E09
3.68E11
8.59E10
Polymers used in
aerospace
equipment and
products
Formulation of
TCEP containing
reactive resin
9.89E06
18,706
N/A
4.62E04
4.41E09
1.03E09
3.46E11
8.07E10
Processing/
Incorporation
into article
Aerospace
equipment and
products and
automotive
articles and
replacement parts
containing TCEP
Processing into 2-
part resin article
N/A
N/A
2.12E06
N/A
5.15E08
1.20E08
5.05E10
1.18E10
Commercial
use/Laboratory
chemicals
Laboratory
chemical
Use of laboratory
chemicals
4.10E09
2.60E06
N/A
5.30E06
4.60E08
1.07E08
4.20E10
9.81E09
Commercial and
Industrial
use/Paints and
coatings
Paints and
coatings
Use of paints and
coatings at job
sites
6.96E07
4.47E04
N/A
8.98E04
2.98E05
6.96E04
5.72E07
1.34E07
Page 311 of 638
-------
Table 5-60. Chronic Fish Ingestion Non-cancer Risk Summary
cou
General Population
Subsistence Fishers4
Tribes (Current)'
Tribes (Heritage)''
Life Cycle
Subcategory
OES
BAF 2,198a
BAF 109a
BAF 2,198
BAF
BAF
BAF
BAF
BAF
Stage/Category
CTe
HE
CTe
HE
109
2,198
109
2,198
109
Manufacturing/
Import
Import
Repackaging
23
5
461
105
1
16
1
11
<1
1
Incorporation into
5
1
104
24
<1
4
<1
2
<1
<1
paints and coatings
Paint and
- 1 -part coatings
Processing/
Incorporation
into formulation,
coating
manufacturing
Incorporation into
paints and coatings
- 2-part reactive
6
1
115
26
<1
4
<1
3
<1
<1
mixture, or
coatings
reaction product
Polymers used
in aerospace
equipment and
products
Formulation of
TCEP containing
reactive resin
4
1
82
19
<1
3
<1
2
<1
<1
Processing/
Aerospace
Processing into 2-
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Incorporation
into article
equipment and
products and
automotive
articles and
replacement
parts containing
TCEP
part resin article
Commercial
Laboratory
Use of laboratory
571
130
11,505
2,617
20
407
13
268
2
35
use/Laboratory
chemicals
chemical
chemicals
Commercial and
Paints and
Use of paints and
10
2
196
45
<1
7
<1
5
<1
1
Industrial
use/Paints and
coatings
coatings at job sites
coatings
" General population exposure estimates based on general population fish ingestion rate of 22.2 g/day.
4 Subsistence fishers exposure estimates based on subsistence fisher ingestion rate of 142.2 g/day.
c Current fish consumption rate at 216 g/day based on survey of Suquamish Indian Tribe in Washington (see Section 5.1.3.4.4).
d Heritage fish consumption rate at 1,646 g/day based on study of Kootenai Tribe in Idaho (see Section 5.1.3.4.4).
* Exposure estimates based on a general population mean fish ingestion rate of 5.04 g/day.
Page 312 of 638
-------
Table 5-61. General Population Lifetime Cancer Oral Ingestion Risk Summary Ta
)le
cou
OES
Lifetime Cancer Oral Risk Estimates
Drinking Water
Drinking Water (Diluted)
Life Cycle Stage/Category
Subcategory
Lifetime
from Birth
Adult
Lifetime
Lifetime from
Birth
Adult Lifetime
Manufacturing/Import
Import
Repackaging
6.09E-07
6.37E-07
3.91E-10
4.09E-10
Processing/Incorporation into
formulation, mixture, or reaction
product
Paint and coating manufacturing
Incorporation into paints and
coatings - 1 -part coatings
2.70E-06
2.82E-06
1.45E-09
1.52E-09
Incorporation into paints and
coatings - 2-part reactive
coatings
2.44E-06
2.56E-06
1.32E-09
1.38E-09
Processing/Incorporation into
formulation, mixture, or reaction
product
Polymers used in aerospace
equipment and products
Formulation of TCEP containing
reactive resin
1.43E-06
1.50E-06
9.20E-10
9.63E-10
Commercial use/Laboratory
chemicals
Laboratory chemical
Use of laboratory chemicals
5.44E-08
5.69E-08
3.49E-11
3.65E-11
Commercial and Industrial
use/Paints and coatings
Paints and coatings
Use of paints and coatings at job
sites
3.42E-06
3.58E-06
6.47E-09
6.77E-09
Page 313 of 638
-------
Table 5-62. Lifetime Cancer Risk Summary for Fish Consumption
cou
OES
Lifetime Cancer Oral Risk Estimates
Life Cycle
Stage/
Category
Subcategory
Adult Fish Ingestion General Population"
Adult Subsistence
Fisher
Tribes
(Current IR)
Tribes
(Heritage IR)
BAF 2,198
BAF 109
BAF
2,198
BAF
109
BAF
2,198
BAF
109
BAF
2,198
BAF
109
CTA
HE
CT"
HE
Manufacturing/
Import
Import
Repackaging
2.32E-03
1.02E-02
1.15E-04
5.07E-04
6.57E-02
3.26E-03
9.96E-02
4.94E-03
7.59E-01
3.77E-02
Processing/
Incorporation
into formulation,
mixture, or
reaction product
Paint and coating
manufacturing
Incorporation into
paints and coatings
- 1 -part coatings
1.03E-02
4.53E-02
5.11E-04
2.25E-03
2.91E-01
1.44E-02
4.41E-01
2.19E-02
3.36
1.67E-01
Incorporation into
paints and coatings
- 2-part reactive
coatings
9.34E-03
4.11E-02
4.63E-04
2.04E-03
2.64E-01
1.31E-02
4.00E-01
1.98E-02
3.05
1.51 E—01
Polymers used in
aerospace
equipment and
products
Formulation of
TCEP containing
reactive resin
1.31E-02
5.74E-02
6.48E-04
2.85E-03
3.69E-01
1.83E-02
5.60E-01
2.78E-02
4.27
2.12E-01
Commercial
use/Laboratory
chemicals
Laboratory
chemical
Use of laboratory
chemicals
9.32E-05
4.10E-04
4.62E-06
2.03E-05
2.63E-03
1.31E-04
3.99E-03
1.98E-04
3.04E-02
1.51E-03
Commercial and
Industrial
use/Paints and
coatings
Paints and
coatings
Use of paints and
coatings at job sites
5.47E-03
2.41E-02
2.71E-04
1.19E-03
1.55E-01
7.67E-03
2.35E-01
1.16E-02
1.79
8.87E-02
" Cancer risk estimates for the adult general population are based on the high-end fish ingestion rate of 22.2 g/day.
4 Exposure estimates are based on a general population mean fish ingestion rate of 5.04 g/day.
Page 314 of 638
-------
Table 5-63. General Population Dermal Acute and Chronic Non-cancer Risk Summary
cou
OES
Acute MOEs
UFs = 30
Chronic Non-cancer MOEa
UFs =30
Life Cycle
Stage/Category
Subcategory
Surface Water
(Adult
Swimming)
Surface Water
(Adult
Swimming)
Child
Playing in
Mud at
100 ma
Child
Activities with
Soil at 100 ma
Child Playing
in Mud at
1,000 ma
Child
Activities with
Soil at 1,000
ma
Manufacturing/Import
Import
Repackaging
6.82E03
4.55E05
6.95E06
1.43E09
5.44E08
1.12E11
Processing/Processing -
Incorporation into
formulation, mixture, or
reaction product
Flame retardant
in: paint and
coating
manufacturing
Incorporation into paints
and coatings - 1 -part
coatings
1.54E03
1.05E05
2.21E05
4.55E07
2.51E07
5.15E09
Incorporation into paints
and coatings - 2-part
reactive coatings
1.70E03
1.14E05
1.53E06
3.14E08
1.16E08
2.39E10
Polymers used in
aerospace
equipment and
products
Formulation of TCEP
containing reactive resin
1.21E03
9.75E04
1.39E06
2.86E08
1.09E08
2.24E10
Processing/Processing -
Incorporation into article
Aerospace
equipment and
products and
automotive
articles and
replacement parts
containing TCEP
Processing into 2-part
resin article
N/A
N/A
1.62E05
3.34E07
1.59E07
3.27E09
Commercial use
Laboratory
chemicals
Use of laboratory
chemicals
1.70E05
1.13E07
1.45E05
2.98E07
1.33E07
2.72E09
Commercial and
Industrial use/Paints and
coatings
Paints and
coatings
Use of paints and
coatings at job sites
2.90E03
1.95E05
9.4E01
1.93E04
1.80E04
3.71E06
" A soil concentration based of annual air deposition fluxes is used to estimate the acute exposures scenario of a child playing with mud and conducting activities in soil.
Page 315 of 638
-------
Table 5-64. Inhalation Chronic Risk Summary for General Population"
cou
OES
Chronic Inhalation MOEs
UFs = 30
Life Cycle Stage/Category
Subcategory
Ambient Air 50th
Ambient Air 95th
Manufacturing/Import
Import
Repackaging
9.34E07
5.10E07
Processing/Incorporation into
formulation, mixture, or
reaction product
Paint and coating
manufacturing
Incorporation into paints and coatings -
1-part coatings
3.66E06
1.49E06
Incorporation into paints and coatings -
2-part reactive coatings
2.22E07
7.18E06
Polymers used in
aerospace equipment and
products
Formulation of TCEP containing reactive
resin
1.98E07
6.41E06
Processing/Incorporation into
article
Aerospace equipment and
products and automotive
articles and replacement
parts containing TCEP
Processing into 2-part resin article
2.41E06
1.82E06
Commercial use/Laboratory
chemicals
Laboratory chemical
Use of laboratory chemicals
2.10E06
1.48E06
Commercial and Industrial
use/Paints and coatings
Paints and coatings
Use of paints and coatings at job sites
1.23E03
4.98E02
" 2,500 lb Production Volume - High-End Release Estimate, Suburban Forest Scenario at 10 m, high meteorology conditions
Page 316 of 638
-------
Table 5-65. General Population Lifetime Cancer Inhalation Risk Summary Table"
cou
OES
Distances
(m)
Lifetime Cancer Inhalation Risk
Life Cycle
Stage/C atego ry
Subcategory
Central Tendency
Meteorological Data
High-End Meteorological Data
Cancer Risk Estimate
for 50th Percentile Air
Concentration
Cancer Risk Estimate
for 95th Percentile Air
Concentration
Cancer Risk Estimate
for 50th Percentile Air
Concentration
Cancer Risk Estimate
for 95th Percentile Air
Concentration
Commercial
and Industrial
use/Paints and
coatings
Paints and
coatings
Use in paints
and coatings
at job sites
10
4.55E-05
5.53E-05
5.42E-05
1.33E-04
30
1.49E-05
2.50E-05
1.50E-05
3.74E-05
30-60
6.74E-06
1.70E-05
7.16E-06
2.29E-05
60
4.63E-06
1.07E-05
4.76E-06
1.20E-05
100
1.66E-06
4.66E-06
1.69E-06
4.46E-06
" 2,500 lb Production Volume - High-End Release Estimate, Suburban Forest Scenario
Page 317 of 638
-------
5.2.8 Hazard Considerations for Aggregate Exposure
For use in this risk evaluation and assessing risks from other exposure routes, EPA conducted route-to-
route extrapolation of the toxicity values from the oral studies for use in the dermal and inhalation
exposure routes and scenarios. Because the health outcomes are systemic and EPA did not identify
evidence of differences in toxicokinetics across exposure routes), EPA considers it is possible to
aggregate risks across exposure routes for all exposure durations and endpoints for the selected PODs
identified in Sections 5.2.5.1 and 5.2.5.2 but only if exposure scenarios indicate that aggregation is
reasonable.
5.2.9 Genotoxicity Hazard Identification and Evidence Integration
For TCEP, several studies evaluated tests of clastogenicity (three in vivo micronucleus assays and one in
vitro chromosomal aberrations assay in mammalian cells), gene mutations (one forward mutation assay
in mammalian cells and six bacterial reverse mutation assays), and other genotoxicity and related
endpoints (two sister chromatid exchange assays, three comet assays, two cell transformation assays,
and one DNA binding assay) specific to TCEP. Although EPA did not evaluate these studies using
formal data quality criteria, selected studies were reviewed by comparing against current OECD test
guidelines and important deviations are noted below. EPA did not review the multiple studies that were
negative for gene mutations. When interpreting the results of these studies, EPA also consulted OECD
(2017V
Tests of clastogenicity and gene mutations can identify the potential for a chemical to induce permanent,
transmissible changes in the amount, chemical properties, or structure of DNA. One of three in vivo
micronucleus assays was readily available. Sala et al. (1982) administered TCEP via i.p. injection to
Chinese hamsters up to 250 mg/kg-day. Study methods deviated from OECD TG 474 (OECD. 2016b) in
several ways. Fewer erythrocytes (2,000 vs. 4,000) were scored than recommended, and the authors did
not verify that TCEP reached the bone marrow, although statistically significant results suggest this was
likely. Sala et al. (1982) used two hamsters per sex vs. five per sex recommended by OECD TG 474 and
used an exposure route that was not recommended. A firm conclusion is not possible given several
deviations from OECD TG 474. Also, the authors state that differences in the response between sexes
with variations among doses make interpretation difficult, resulting in an equivocal conclusion.
However, EPA combined results across sexes, based on a comparison of means test that indicated
similar results across sex and dose. This allowed greater statistical power (OECD. 2017). These combined
results showed statistically significant increases in micronuclei that showed a dose-response trend. No
information was provided to allow comparison with historical controls.
Two negative in vivo micronucleus studies using mice cited in the 2009 European Union Risk
Assessment Report (ECB. 2009) and a review article (Beth-Hubner. 1999) were not available for review.42
TCEP also did not induce chromosomal aberrations in an in vitro assay using Chinese hamster ovary
cells (Galloway et al.. 1987) that was mostly consistent with OECD TG 473 (OECD. 2016a). except that
the authors scored only 100 cells per concentration compared with the recommended 300 per
concentration needed to conclude that a test is clearly negative.
A forward gene mutation assay using Chinese hamster lung fibroblasts (Sala et al.. 1982) and multiple
bacterial reverse gene mutation assays (Follmann and Wober. 2006; Haworth et al.. 1983; BIBRA. 1977;
42 According to ECB (2009). the mouse i.p. study used doses from 175 to 700 mg/kg-day, and the oral study used a dose of
1,000 mg/kg. The original reports were not readily available for review.
Page 318 of 638
-------
Prival et al.. 1977; Simmon et al.. 1977) were all negative for the induction of gene mutations. Most in
vitro gene mutation assays were conducted both with and without metabolic activation. In a study by
Nakamura et al. (1979). TCEP induced gene mutations in two Salmonella typhimurium strains. In strain
TA1535, increases of four to seven times the control response were observed only with metabolic
activation and in TA100, increases were observed both with and without metabolic activation. The
reason for the inconsistency in results between Nakamura et al. (1979) and the other studies is unclear
because concentrations were comparable. One difference, however, is that Nakamura et al. (1979) used a
mixture of PCBs (Kanechlor 500) for metabolic activation, whereas other studies used Aroclor 1254 or
did not appear to induce enzymes in the S9 fractions.
In addition to clastogenicity and gene mutation tests, other genotoxicity tests that measured DNA
damage or DNA binding been conducted using TCEP. Two sister chromatid exchange (SCE) assays
identified (1) equivocal results in Chinese hamster ovary cells (Galloway et al.. 1987). and (2)
statistically significant differences from controls in Chinese hamster lung fibroblasts but no clear dose
response (Sala et al.. 1982). In vitro comet assays in peripheral mononuclear blood cells (PMBCs)
identified DNA damage at the highest concentration, although it is not known whether this result was in
the presence of cytotoxicity (Bukowski et al.. 2019). Another comet assay did not identify DNA damage
in Chinese hamster fibroblasts either with or without metabolic activation (Follmann and Wober. 2006).
TCEP was also negative in a DNA binding assay (Lown et al.. 1980).
Sala et al. (1982) identified a high level of cell transformation in Syrian hamster embryo (SHE) cells but
a lower level using C3H10T1/2 cells with metabolic activation. These cell transformation results may
reflect direct or indirect genetic interactions or non-genotoxic mechanisms (OECD. 2007).
Overall, direct mutagenicity is not expected to be a predominant mode of action. Appendix M provides
additional details regarding TCEP genotoxicity studies as well as considerations regarding the quality of
the studies.
U.S. EPA's PPRTV (U.S. EPA. 2009) concluded that the overall weight of scientific evidence for the
mutagenicity of TCEP is negative. The PPRTV also acknowledged the weak positive result in the Ames
assay by Nakamura et al. (1979) and characterized the in vivo micronucleus assay in Chinese hamsters
(Sala et al.. 1982) as equivocal.
Page 319 of 638
-------
5.3 Human Health Risk Characterization
Human Health Risk Characterization (Section 5.3):
Key Points
EPA evaluated all reasonably available information to support human health risk characterization.
The key points of the human health risk characterization are summarized below:
• Dermal exposures drive risks to workers in occupational settings and both cancer risks and
non-cancer MOEs that met benchmarks were observed for most COUs, whereas risks and
MOEs from inhalation exposure met benchmarks for multiple commercial paints and
coatings use scenarios within a single COU.
• Fish ingestion is the primary exposure route driving risks to the general population. People
who are subsistence fishers may be at high risk if they eat TCEP-contaminated fish; tribal
people for whom fish is important dietarily and culturally may have higher risks than the
general population and subsistence fishers.
• Mouthing by infants and children is the primary exposure risk for consumer articles to which
infants and children are exposed.
• Infants exposed through human milk ingestion are not more sensitive than the mothers. The
COUs that present infant risks also result in maternal risks. There are no COUs that show
infant risks but not maternal risks. Therefore, protecting the mother will also protect the
infant from exposure via human milk.
5.3.1 Risk Characterization Approach
The exposure scenarios, populations of interest, and toxicological endpoints used for evaluating risks
from acute, intermediate, and chronic/lifetime exposures are summarized in Table 5-66.
Table 5-66. Exposure Scenarios, Populations of Interest, and Hazard Values
Populations of Interest
and Exposure Scenarios
Workers
Male and female adolescents and adults (>16 years old) directly working with TCEP
under light activity (breathing rate of 1.25 m3/hr)
Exposure durations
• Acute - 8 hours for a single workday (most OESs)
• Intermediate - 8 hours per workday for 22 working days
• Chronic - 8 hours per workday for 250 days per year for 31 or 40 working
years
Exposure routes - Inhalation and dermal
Occupational Non-users
Male and female adolescents and adults (>16 years old) indirectly exposed to TCEP
within the same work area as workers (breathing rate of 1.25 m3/hr)
Exposure durations
• Acute, Intermediate, and Chronic - same as workers
Exposure route - Inhalation
Consumers
Male and female infants, children and adults using articles that contains TCEP
Exposure durations
• Acute - 1 day exposure
Page 320 of 638
-------
• Chronic - 365 days per year
Exposure routes
• Adults - Inhalation and dermal
• Infants and Children - Inhalation, dermal, and oral
Populations of Interest
and Exposure Scenarios
General Population
Male and female infants, children and adults exposed to TCEP through drinking
water, ambient water, ambient air, soil, and diet
Exposure durations
• Acute - Exposed to TCEP continuously for a 24-hour period
• Chronic - Exposed to TCEP continuously up to 33 years
Exposure routes - Inhalation, dermal, and oral (depending on exposure scenario)
Infants (Human Milk Pathway)
Infants exposed to TCEP through human milk ingestion
Exposure durations
• Intermediate - Exposed to TCEP continuously for 30 days
• Chronic - Exposed to TCEP continuously for one year
Exposure routes - Oral
Non-cancer Acute Hazard Values6
Sensitive health effect: Neurotoxicity
HECdi7,7v, continuous = 51.5 mg/m3 (4.41 ppm)
HEDdi7,/v = 9.46 mg/kg; dermal and oral
Total acute UF (benchmark MOE) = 30 (UFa= 3; UFh= 10)c
Health Effects, Hazard
Values, and Benchmarks
Non-cancer Intermediate/Chronic Values6
Sensitive health effect: Male reproductive effects
HECdi7,7v, continuous = 14.9 mg/m3 (1.27 ppm)
HED/A„/r = 2.73 mg/kg; dermal and oral
Total intermediate/chronic UFs (benchmark MOE) = 30 (UFA= 3; UFH= 10)c
Cancer Hazard Values6
Both values based on renal tumors
IURd«7v, continuous = 0.00451 per mg/m3 (0.0526 per ppm)
CSF Daily = 0.0245 per mg/kg-day
11 The chronic duration is the most relevant exposure scenario for the consumer COUs and is used to assess chronic
non-cancer and lifetime cancer risks. Acute exposure duration non-cancer risks are presented to help characterize risk.
h The inhalation HEC and IUR are extrapolated from the oral HED or CSF, which are estimated using allometric scaling
(BW3 4) and are associated with continuous or daily exposures. The HEC and IUR values assume a resting breathing
rate (0.6125 m3/hr). The dermal HED is assumed to equal the oral HED. See Appendix K.3 and Benchmark Dose
Modeling Results for TCEP in U.S. EPA (2024c) for dose derivation.
c Total UFs in the benchmark MOE.
UFa = interspecies (animal to human); UFh = intraspecies (human variability)
5.3.1.1 Estimation of Non-cancer Risks
EPA used a margin of exposure (MOE) approach to identify potential non-cancer risks. The MOE is the
ratio of the non-cancer POD divided by a human exposure dose. Acute, intermediate, and chronic MOEs
for non-cancer inhalation and dermal risks were calculated using the following equation:
Equation 5-26.
Non — Cancer Hazard Value (POD)
M0E= Human Exposure
Page 321 of 638
-------
Where:
MOE
Non-cancer Hazard Value (POD) =
Human Exposure =
Margin of exposure for acute, intermediate, or
chronic risk comparison (unitless)
HEC (mg/m3) or HED (mg/kg-day)
Exposure estimate (mg/m3 or mg/kg-day)
MOE risk estimates may be interpreted in relation to benchmark MOEs. Benchmark MOEs are typically
the total UF for each non-cancer POD. The MOE estimate is interpreted as a human health risk of
concern if the MOE estimate is less than the benchmark MOE (i.e., the total UF). On the other hand, if
the MOE estimate is equal to or exceeds the benchmark MOE, the risk is not considered to be of concern
and mitigation is not needed. Typically, the larger the MOE, the more unlikely it is that a non-cancer
adverse effect occurs relative to the benchmark. When determining whether a chemical substance
presents unreasonable risk to human health or the environment, calculated risk estimates are not "bright-
line" indicators of unreasonable risk, and EPA has the discretion to consider other risk-related factors in
addition to risks identified in the risk characterization.
5.3.1.2 Estimation of Cancer Risks
Extra cancer risks for repeated exposures to a chemical were estimated using the following equations:
Equation 5-27.
Inhalation Cancer Risk = Human Exposure x IUR
or
Dermal or Oral Cancer Risk = Human Exposure x CSF
Where:
Risk = Extra cancer risk (unitless)
Human Exposure = Exposure estimate (LADC in ppm)
IUR = Inhalation unit risk (risk per mg/m3)
CSF = Cancer slope factor (risk per mg/kg-day)
Estimates of extra cancer risks are interpreted as the incremental probability of an individual developing
cancer over a lifetime following exposure (i.e., incremental or extra individual lifetime cancer risk).
EPA considers a range of extra cancer risk from 1 x 1CT4 to 1 x 10 6 to be relevant benchmarks for risk
assessment (U.S. EPA 2017b). Consistent with NIOSH guidance (Whittaker et al.. 2016). under TSCA,
EPA typically applies a 1 x 10~4 benchmark for occupational scenarios in industrial and commercial work
environments subject to OSHA requirements. The Agency typically considers the general population
and consumer benchmark for cancer risk to be within the range of 1x 10~6 and 1 x 10 4. Again, it is
important to note that these benchmarks are not bright lines and EPA has discretion to find unreasonable
risks based on other risk-related considerations based on analysis. Exposure-related considerations (e.g.,
duration, magnitude, population exposed) can affect EPA's estimates of the excess lifetime cancer risk.
5.3.2 Summary of Human Health Risk Characterization
5.3.2.1 Summary of Risk Estimates for Workers
EPA estimated cancer risks and non-cancer MOEs for workers exposed to TCEP for multiple COUs
based on the occupational exposure estimates described in Section 5.3.2.1.1. Complete risk calculations
and results for the occupational OES/COUs are available in Risk Evaluation for Tris(2-chloroethyl)
Page 322 of 638
-------
Phosphate (TCEP) - Supplemental Information File: Risk Calculator for Occupational Exposures (U.S.
EPA. 2024m).
5.3.2.1.1 COUs/OESs with Quantitative Risk Estimates
Table 5-67. summarizes cancer and non-cancer risk estimates for the inhalation and dermal exposures
for all OESs assessed. These risk estimates are based on exposures estimated for workers who do not use
PPE such as gloves or respirators. When both monitoring and modeling data were available for
inhalation exposures, EPA only presented the risk estimates for the most reliable data source in the
summary table. Estimates for inhalation and dermal exposures that have PPE factored in are contained in
the Risk Evaluation for TCEP - Supplemental Information File: Risk Calculator for Occupational
Exposures (U.S. EPA. 2024m).
Exposure data for ONUs were not available for most COUs except for recycling (with recycling e-waste
as the relevant OES). For the COUs and OESs without ONU-specific exposure data, EPA assumed risks
would be equal to or less than risks to workers who handle materials containing TCEP as part of their
job. The inhalation risk values used for workers are also presented for ONUs in Table 5-67.. EPA
assumed that ONUs are not exposed dermally.
Within the commercial use of paints and coatings COU, EPA did not calculate short-term or chronic,
non-cancer risks or lifetime cancer risks for the 1-day spray application for commercial paint and
coating scenarios (OES #7 and #10) because risks were most appropriately assessed using only the
inhalation HEC and dermal HED values for acute exposures. Likewise, EPA did not calculate chronic
non-cancer or lifetime cancer risks for the 2-day commercial paint and coating spray application (OES
#8 and #11) given the very limited number of days per year of exposure. However, for OESs exposures
longer than 1 day per year, EPA also compared exposure with the acute hazard PODs.
EPA was informed by the Auto Alliance that it is possible that TCEP containing articles, replacement
parts, and paints could be in use in the automotive industry, however, no data regarding product(s),
operating site(s), etc was provided. As reflected below, in Table 5-58, in the absence of reasonably
available data to refine the modelling, EPA expects modelled environmental releases and occupational
exposures to be similar to the OES's already previously assessed for other industry sectors on a per
generic site basis. Specifically, this means that that the recently added industrial use of paints and
coatings would be similar to the commercial use of paints and coatings and that the recently modified
COUs of Incorporation into article and Installation of article would be similar regardless of whether or
not it is being done for aircraft or automobiles.
Risks from Inhalation Exposure
Cancer inhalation risk estimates were above 1 in 10,000 for the commercial use of paints and coatings
COU for both central tendency and high-end exposures. These risks were associated with two OESs:
250-day applications of either 1- or 2-part sprays. Risk estimates were less than 1 in 10,000 for the
remaining six occupational COU subcategories.
In addition, inhalation non-cancer MOEs were less than benchmark MOEs for the commercial use of
paints and coatings COU for high-end exposures. Within this COU, high-end acute exposure for all
three OESs associated with 2-part spray applications resulted in MOEs less than the benchmark MOE of
30. For high-end short-term/chronic exposures, MOEs were less than the benchmark MOE of 30 for the
250-day applications of either 1- or 2-part sprays. No other COU/OES combinations resulted in MOEs
less than the non-cancer benchmark MOEs; this includes the commercial and industrial uses for the
Page 323 of 638
-------
installation of articles, which used surrogate monitoring data to estimate inhalation exposures that could
occur during these activities.
Risks from Dermal Exposure
More COU categories were associated with worker dermal risks above 1 in 10,000. Cancer dermal risk
estimates were above 1 in 10,000 for both central tendency and high-end exposures for certain
subcategories and OESs within the following five COU categories: (1) Import; (2) Incorporation into
formulation, mixture, or reaction products; (3) Processing - Incorporation into an article; (4)
Commercial use of paints and coatings; and (5) Other commercial use - Laboratory chemicals.
Additional dermal cancer risks above 1 in 10,000 were observed for only high-end exposures within a
single COU category (Processing - Incorporation into formulations, mixtures, or reaction products) and
two associated OESs (Incorporation into 2-part paints and coatings and Formulation of 2-part reactive
resins).
Three COU categories had chronic non-cancer dermal MOEs less than the benchmark value of 30 for
both high-end and central tendency exposures. These were, (1) Processing - Incorporation into articles;
(2) Commercial use of paints and coatings; and (3) Other commercial use - Laboratory chemicals. Two
additional COUs were associated with MOEs lower than 30 for only high-end exposures, these were (1)
Import; and (2) Processing - Incorporation into formulation, mixture, or reaction products.
For the short-term exposure scenario, MOEs were less than 30 for five COUs for at least some OESs.
Within two of these COUs, certain OESs had MOEs less than 30 for only high-end exposures—
Flame retardant in paints and coatings manufacture (2-part coatings and polymers in aerospace
equipment) and Commercial use of paints and coatings (2-day application for 1-part coatings).
For the acute exposure scenario, five COUs had dermal MOEs of less than 30 for both central tendency
and high-end exposures. One of these five COUs (commercial use of paints and coatings) also had some
OESs (1-part sprays) for which MOEs were less than 30 for only high-end exposures.
Processing/Recycling was the single COU with cancer dermal risks less than 1 in 10,000 and all non-
cancer MOEs greater than benchmark values. Dermal risk estimates were not calculated for industrial
and commercial use of aerospace equipment and automotive products because EPA does not expect
dermal exposure for this COU because TCEP will be entrained in the polymer matrix.
Page 324 of 638
-------
Table 5-67. Occupational Risk Summary for 2,500 lb Production Volume
cou
OES
Population
Exposure
Route and
Duration
Exposure
Level
Estimates for No PPE
Overall
Confidence
in Risk
Estimates
Life Cycle
Stage/
Category
Subcategory
Acute Non-
cancer MOE
UFs = 30
Intermediate Non-
cancer MOE
UFs =30
Chronic Non-
cancer MOE
UFs =30
Lifetime
Cancer
Risk
Manufacturing
/ Import
Import
Repackaging
Worker
Inhalation
8-hr TWA
CT
6.8E03
1.4E04
1.7E05
1.5E-07
Moderate
High-End
1.9E03
4.0E03
4.9E04
5.5E-07
ONU"
Inhalation
8-hr TWA
CT
6.8E03
1.4E04
1.7E05
1.5E-07
Slight
High-End
1.9E03
4.0E03
4.9E04
5.5E-07
Worker
Dermal
CT
4.3
9.4
1.14E02
2.3E-04
Moderate
High-End
1.4
1.8
2.2E01
1.6E-03
Processing/
Processing -
Incorporation
into
formulation,
mixture, or
reaction
product
Flame retardant in:
Paint and coating
manufacturing
Incorporation
into paints and
coatings - 1 -
part coatings
Worker
Inhalation
8-hr TWA
CT
4.6E03
6.7E03
7.7E04
3.3E-07
Moderate
High-End
7.3E02
1.6E03
1.9E04
1.4E-06
ONU"
Inhalation
8-hr TWA
CT
4.6E03
6.7E03
7.7E04
3.3E-07
Slight
High-End
7.3E02
1.6E03
1.9E04
1.4E-06
Worker
Dermal
CT
4.3
6.3
7.6E01
3.5E-04
Moderate
High-End
1.4
5.7E-01
4.0
8.6E-03
Incorporation
into paints and
coatings - 2-
part coatings
Worker
Inhalation
8-hr TWA
CT
7.9E02
6.5E03
7.9E04
3.2E-07
Moderate
High-End
1.9E02
1.6E03
1.9E04
1.4E-06
ONU"
Inhalation
8-hr TWA
CT
7.9E02
6.5E03
7.9E04
3.2E-07
Slight
High-End
1.9E02
1.6E03
1.9E04
1.4E-06
Worker
Dermal
CT
4.3
3.8E01
4.6E02
5.8E-05
Moderate
High-End
1.4
6.3
7.6E01
4.5E-04
Polymers used in
aerospace equipment
and products
Formulation of
TCEP into 2-
part reactive
resin
Worker
Inhalation
8-hr TWA
CT
1.0E04
6.7E03
8.1E04
3.1E-07
Moderate
High-End
1.9E02
1.5E03
1.8E04
1.5E-06
ONU"
Inhalation
8-hr TWA
CT
1.0E04
6.7E03
8.1E04
3.1E-07
Slight
High-End
1.9E02
1.5E03
1.8E04
1.5E-06
Worker
Dermal
CT
4.3
3.8E01
4.6E02
5.8E-05
Moderate
High-End
1.4
2.1
2.5E01
1.4E-03
Processing/
Processing -
Incorporation
into article
Aerospace equipment
and products and
automotive articles and
replacement parts
containing TCEP
Processing into
2-part resin
article
Worker
Inhalation
8-hr TWA
CT
2.2E04
9.0E03
3.8E04
6.6E-07
Moderate
High-End
4.2E03
1.8E03
6.3E03
4.1E-06
ONU"
Inhalation
8-hr TWA
CT
2.2E04
9.0E03
3.8E04
6.6E-07
Slight
High-End
4.2E03
1.8E03
6.3E03
4.1E-06
Worker
Dermal
CT
1.1E01
4.3
1.6E01
1.7E-03
Moderate
High-End
3.6
1.4
1.5
2.3E-02
Page 325 of 638
-------
cou
Exposure
Route and
Duration
Estimates for No PPE
Overall
Life Cycle
Stage/
Category
Subcategory
OES
Population
Exposure
Level
Acute Non-
cancer MOE
UFs = 30
Intermediate Non-
cancer MOE
UFs =30
Chronic Non-
cancer MOE
UFs =30
Lifetime
Cancer
Risk
Confidence
in Risk
Estimates
Worker
Inhalation
CT
7.6E08
3.0E08
3.2E08
8.4E-11
Moderate
Processing -
Recycling e-
waste
8-hr TWA
High-End
7.8E04
3.1E04
3.3E04
1.0E-06
Processing/
Recycling
ONU
Inhalation
CT
7.6E08
3.0E08
3.2E08
8.4E-11
Moderate
Recycling
8-hr TWA
High-End
4.0E05
1.6E05
1.7E05
2.0E-07
Worker
Dermal
CT
5.2E05
2.0E05
2.2E05
1.2E-07
Moderate
High-End
2.2E05
8.5E4
9.1E04
3.8E-07
Worker
Inhalation
CT
4.5E02
N/A
N/A
N/A
Moderate
Paints and
8-hr TWA
High-End
6.9E01
N/A
N/A
N/A
coatings -
Spray (1-part
coatings, 1-day
application)
ONU"
Inhalation
CT
4.5E02
N/A
N/A
N/A
Slight
8-hr TWA
High-End
6.9E01
N/A
N/A
N/A
Worker
Dermal
CT
3.2E01
N/A
N/A
N/A
Moderate
High-End
5.9
N/A
N/A
N/A
Worker
Inhalation
CT
4.5E02
1.9E03
N/A
N/A
Moderate
Paints and
8-hr TWA
High-End
6.9E01
3.0E02
N/A
N/A
coatings -
Spray (1-part
coatings, 2-day
application)
ONU"
Inhalation
CT
4.5E02
1.9E03
N/A
N/A
Slight
8-hr TWA
High-End
6.9E01
3.0E02
N/A
N/A
Worker
Dermal
CT
3.2E01
1.4E02
N/A
N/A
Moderate
High-End
5.9
2.6E01
N/A
N/A
Commercial
Paints and coatings
Paints and
coatings -
Spray (1-part
Worker
Inhalation
CT
4.5E02
1.8E02
1.9E02
1.4E-04
Moderate
and Industrial
use/Paints and
coatings
8-hr TWA
High-End
6.9E01
2.7E01
2.9E01
1.2E-03
ONU"
Inhalation
CT
4.5E02
1.8E02
1.9E02
1.4E-04
Slight
coatings, 250-
8-hr TWA
High-End
6.9E01
2.7E01
2.9E01
1.2E-03
day
application)
Worker
Dermal
CT
3.2E01
1.3E01
1.3E01
2.0E-03
Moderate
High-End
5.9
2.3
2.5
1.4E-02
Worker
Inhalation
CT
9.0E01
N/A
N/A
N/A
Moderate
Paints and
8-hr TWA
High-End
1.4E01
N/A
N/A
N/A
coatings -
Spray (2-part
coatings, 1-day
application)
ONU"
Inhalation
CT
9.0E01
N/A
N/A
N/A
Slight
8-hr TWA
High-End
1.4E01
N/A
N/A
N/A
Worker
Dermal
CT
6.4
N/A
N/A
N/A
Moderate
High-End
1.2
N/A
N/A
N/A
Paints and
coatings -
Spray (2-part
Worker
Inhalation
CT
9.0E01
3.9E02
N/A
N/A
Moderate
8-hr TWA
High-End
1.4E01
5.9E01
N/A
N/A
ONU"
Inhalation
CT
9.0E01
3.9E02
N/A
N/A
Slight
8-hr TWA
High-End
1.4E01
5.9E01
N/A
N/A
Page 326 of 638
-------
cou
OES
coatings, 2-day
application)
Population
Exposure
Route and
Duration
Exposure
Level
Estimates for No PPE
Overall
Confidence
in Risk
Estimates
Life Cycle
Stage/
Category
Subcategory
Acute Non-
cancer MOE
UFs = 30
Intermediate Non-
cancer MOE
UFs =30
Chronic Non-
cancer MOE
UFs =30
Lifetime
Cancer
Risk
Commercial
and Industrial
use/Paints and
coatings
Paints and coatings
Worker
Dermal
CT
6.4
2.8E01
N/A
N/A
Moderate
High-End
1.2
5.1
N/A
N/A
Paints and
coatings -
Spray (2-part
coatings, 250-
day
application)
Worker
Inhalation
8-hr TWA
CT
9.0E01
3.8E01
3.8E01
7.1E-04
Moderate
High-End
1.4E01
5.4
5.8
6.0E-03
ONU"
Inhalation
8-hr TWA
CT
9.0E01
3.8E01
3.8E01
7.1E-04
Slight
High-End
1.4E01
5.4
5.8
6.0E-03
Worker
Dermal
CT
6.4
2.5
2.7
9.9E-03
Moderate
High-End
1.2
4.6E-01
5.0E-01
6.9E-02
Industrial
Use/Other Use
Aerospace equipment
products and
automotive articles and
replacement parts
containing TCEP
Installation of
articles
Worker
Inhalation
8-hr TWA
CT
5.8E06
2.3E06
2.5E06
1.1E-08
Slight
High-End
5.8E06
2.3E06
2.5E06
1.1E-08
ONU"
Inhalation
8-hr TWA
CT
5.8E06
2.3E06
2.5E06
1.1E-08
Slight
High-End
5.8E06
2.3E06
2.5E06
1.1E-08
Worker
Dermal
CT
N/A
N/A
N/A
N/A
N/A
High-End
N/A
N/A
N/A
N/A
Commercial
Use/Other Use
Aerospace equipment
products and
automotive articles and
replacement parts
containing TCEP
Use and/or
maintenance of
articles
Worker
Inhalation
8-hr TWA
CT
5.8E06
2.3E06
2.5E06
1.1E-08
Slight
High-End
5.8E06
2.3E06
2.5E06
1.1E-08
ONU"
Inhalation
8-hr TWA
CT
5.8E06
2.3E06
2.5E06
1.1E-08
Slight
High-End
5.8E06
2.3E06
2.5E06
1.1E-08
Worker
Dermal
CT
N/A
N/A
N/A
N/A
N/A
High-End
N/A
N/A
N/A
N/A
Commercial
Use/Other Use
Laboratory chemicals
Laboratory
chemicals
Worker
Inhalation
8-hr TWA
CT
1.0E05
5.1E04
5.5E04
4.0E-07
Moderate
High-End
6.5E04
3.2E04
3.5E04
6.8E-07
ONU"
Inhalation
8-hr TWA
CT
1.0E05
5.1E04
5.5E04
4.0E-07
Slight
High-End
6.5E04
3.2E04
3.5E04
6.8E-07
Worker
Dermal
CT
4.3
1.7
2.7
9.7E-03
Moderate
High-End
1.4
5.7E-01
7.6E-01
4.5E-02
Disposal/
Disposal
Disposal
Disposal
Evaluated as part of each OES as opposed to a standalone OES
CT = Central Tendency
Page 327 of 638
-------
5.3.2.1.2 COUs/OESs Without Quantitative Risk Estimates
Distribution in Commerce
For purposes of assessment in this risk evaluation, distribution in commerce of TCEP consists of the
transportation associated with the moving of sealed containers of TCEP. EPA expects TCEP to be
transported in sealed containers from import sites to downstream processing and use sites, or for final
disposal of TCEP. The steps of loading and unloading that are assessed during other COUs/OESs
consists of unloading TCEP into the formulation process and loading refers to packaging the finished
product prior to shipment. Loading and unloading activities that occur during a distribution in commerce
scenario would only refer to loading or unloading sealed containers from a transport vehicle. More
broadly under TSCA, "distribution in commerce" and "distribute in commerce" are defined under TSCA
section 3(5). EPA expects under standard operating procedures that exposures and releases that could
occur during distribution in commerce are expected to be negligible because containers remain sealed
during transport.
Commercial Uses that TCEP is No Longer Actively Incorporated Into
EPA determined that the following commercial use COUs for TCEP consists of use of existing products
and end of service life disposal:
• Commercial use - Furnishing, cleaning, treatment/care products - Fabric and textile products;
• Commercial use - Furnishing, cleaning, treatment/care products - Foam seating and bedding
products;
• Commercial use - Construction, paint, electrical, and metal products - Building/construction
materials - Insulation; and
• Commercial use - construction, paint, electrical, and metal products - Building/construction
materials - Wood and engineered wood products - Wood resin composites.
TCEP was used for these purposes in the past, but the COUs were phased out beginning in the late
1980s or early 1990s and replaced by other flame retardants or flame-retardant formulations. EPA did
not locate data to estimate (1) the amount of TCEP used in these products, (2) the amounts of these
products that have already reached the end of their service life, or (3) the amounts that have already been
disposed. Based on the years that the phase-out occurred, many of these products are likely to no longer
be in use because the end of their service life was already reached (e.g., commercial roofing has an
estimated lifespan of 17-20 years). EPA assumes that any of these products still used commercially
represent a fraction of the overall amount of TCEP previously used for these purposes.
EPA acknowledges that workers may handle fabric and textile products, foam seating and bedding
products, building/construction materials - wood and engineered wood products - wood resin
composites, building/construction materials insulation that were previously manufactured when TCEP
was more often present as a flame retardant in these articles. There exists some potential for exposures
to workers in commercial settings that handle or repair older articles that used TCEP.
In addition, office workers may be exposed to fabric and textile products, foam seating and bedding
products, building/construction materials - wood and engineered wood products - wood resin
composites, building/construction materials insulation, in the commercial environments and offices.
Office workers exposed to such articles would mirror the consumer assessment described in Section
5.1.2 due to the similarity of exposure scenarios between office workers and consumers. The consumer
assessment for these articles resulted in no consumer risk for inhalation, ingestion, or dermal risk for
adults for the COUs with moderate confidence. Therefore, EPA has moderate confidence that under
similar exposure durations and exposure frequencies, the Agency expects these articles to pose no
Page 328 of 638
-------
commercial risk for inhalation, ingestion, and dermal risk to COUs to commercial workers who use
articles in a similar fashion to consumers.
Commercial uses may be higher or lower than consumer exposures for a multitude of reasons.
Commercial uses for PESS such as truck drivers (auto-foam), gym teachers (foam mats, and wood
flooring), and DIY hobbyists (wood resin products) may have elevated exposures to these finished
articles due to their increased activity and use patterns with such articles. Whereas commercial
environments may frequently replace older furniture articles with newer articles which no longer include
TCEP.
Due to the lack of reasonably available information on (1) the amount of TCEP used in these products;
(2) the amounts of these products that have already reached the end of their service life; or (3) the
amount of articles that remain in commercial environments; EPA is unable to quantify the exposure to
commercial COUs listed above.
Disposal
Waste handling, disposal, and/or treatment includes waste disposal (landfilling or incineration) as well
as water (e.g., releases to wastewater treatment and POTWs) and air releases (e.g., fugitive and stack air
emissions). Workers engaged in these activities at the facilities where TCEP is processed and used, as
well as workers at off-site waste treatment and disposal facilities (e.g., landfills, incinerators, POTWs)
could be exposed to TCEP.
EPA estimated releases to landfills for the following two COU/OES combinations:
• Processing - Incorporation into formulation, mixture, or reaction product - Paint/coating
manufacture - 1-part coating OES; and
• Processing - Incorporation into articles - Aerospace equipment and products and automotive
articles and replacement parts containing TCEP - Processing in two-part resin article OES.
EPA estimated releases to incinerators for the following two COU/OES combinations:
• Processing - Incorporation into formulation, mixture, or reaction product - Paint/coating
manufacture - 2-part coating OES; and
• Processing - Incorporation into formulation, mixture, or reaction product - Polymers in
aerospace equipment and products - Formulation of reactive resins OES.
Both releases to landfills and incinerators rely on inputs provided by ESDs or GSs. However, the ESDs
and GSs do not specify the proportion of the throughput that goes to either of these two disposal
practices. Therefore, EPA was unable to further quantify environmental releases related to these two
disposal processes.
For three of the COUs/OESs listed above, EPA was able to perform quantitative risk characterization
that included releases to on-site wastewater treatment or discharge to POTWs, where applicable (see
Table 3-2). Any worker exposures associated with on-site waste treatment were combined with other
exposures as relevant for the above COUs.
Waste treatment or disposal is expected to be negligible for industrial and commercial uses related to
installing articles for aerospace applications and automotive articles and replacement parts containing
TCEP. For the COUs of manufacturing/repackaging, commercial use of paints and coatings, commercial
use of laboratory chemicals, and disposal to landfills or incinerators are not expected but EPA estimated
surface water releases that could include release to wastewater treatment or POTWs.
Page 329 of 638
-------
For the commercial uses that have been phased out, any currently used products that contain TCEP are
expected to be disposed in landfills but will represent just a fraction of previous amounts from when
TCEP was used more widely. Data are lacking with which to estimate exposure and risk from disposal
or waste treatment activities for these COUs and EPA has not quantified such risks. For e-waste
recycling, there is also too little information to estimate exposure from disposal and only a small portion
of e-waste is expected to contain TCEP. Therefore, EPA's confidence in these exposures is
indeterminate and cannot quantify risk for the disposal or waste treatment activities for these COUs.
However, EPA acknowledges that while some releases and exposures could occur during the disposal of
the wide variety of items that TCEP has found its way into, these exposures are expected to be
negligible.
5.3.2.2 Summary of Risk Estimates for Consumers
5.3.2.2.1 COUs with Quantitative Risk Estimates
Table 5-68 summarizes the dermal, inhalation, and ingestion MOEs used to characterize non-cancer risk
for acute, intermediate, and chronic exposure and presents these values for all life stages for each COU.
Table 5-69 summarizes the dermal, inhalation, and oral lifetime cancer risk estimates for each consumer
COU. Risk estimates in Table 5-68 and Table 5-69 are only presented for COUs, routes, and age groups
that are below the non-cancer risk benchmarks or above the lifetime cancer benchmarks. For cancer,
EPA uses a cancer benchmark range from 1 in 10,000 to 1 in 1,000,000 to consider and characterize
lifetime cancer risks from consumer exposure. Table 5-69 presents the risk estimates that were above the
lifetime cancer benchmark of 1 in 1,000,000.
Although CEM 3.2 provides inhalation exposure doses for each age group, inhalation exposure risk
estimates were calculated for the adult exposure scenario. Inhalation risk estimates for other life stages
are presented in Appendix J. These adjusted inhalation exposure doses are estimated using breathing rate
and body weight considerations for each age group. Body weight- and inhalation rate-adjusted inhalation
risk estimates for younger life stages should be interpreted with caution. Despite accounting for
breathing rate and body weight, adjusted inhalation exposures for younger age groups may be inaccurate
because there are other considerations (e.g., elimination kinetics) that may differ among age groups
(U.S. EPA. 2012a). Information on the inputs used for consumer modeling using CEM 3.2 are presented
in Section 5.1.2 and Appendix J.
Acute and Chronic Risks
Children and infants have acute oral MOEs less than the benchmark of 30 for foam toy blocks, roofing
insulation, and wood flooring. Infants have acute oral MOEs less than the benchmark of 30 for all of the
COUs except acoustic ceilings. Chronic oral MOEs for children and infants are below the benchmark of
30 for fabric and textiles, foam seating and bedding products, wood articles (e.g., wood flooring and
wooden TV stands). Infants and children have a greater susceptibility to TCEP exposure due to
mouthing behaviors associated with toys (e.g., outdoor play structures, foam blocks). As discussed in
Section 5.1.2.2.4, EPA selected a high mouthing parameter (50 cm2) for the COUs that were designed
for children. For other products that had the potential for mouthing, EPA selected medium mouthing
parameters (10 cm2). Mouthing duration had a pronounced impact on the oral exposures for children and
infants (see Appendix J).
Section 5.1.2.2.3 describes the parameters selection and assumptions considered for the dermal exposure
assessment. Acute and chronic dermal MOEs for all life stages are below the benchmark of 30 for wood
flooring. Chronic dermal MOEs for children and infants are below the benchmark of 30 for wood
articles (e.g., wood flooring and wooden TV stands). Sensitivity analyses indicated that the initial SVOC
Page 330 of 638
-------
concentration in the article (a product of the article density and the weight fraction) is a driver of dermal
exposures. The consumer modeling suggests direct contact with wooden articles (e.g., wood flooring,
wooden TV stands) results in greater exposure than dermal doses mediated from dust generated from
consumer articles.
Chronic inhalation MOEs for acoustic ceilings, wood flooring, and insulation are below the benchmark
of 30. Acute inhalation MOEs for textiles in outdoor play structures, acoustic ceilings, wood flooring,
wooden TV stands, and insulation are below the benchmark of 30. Furthermore, figuresFigure
5-20,Figure 5-21, andFigure 5-22 demonstrate the chronic exposure estimates may be high due to an
initial off-gassing period where there are high gas phase air concentrations. Depending on the COU,
after a few weeks to a few months, there is a precipitous drop in the gas phase air concentrations. These
findings suggest that inhalation exposure estimates are higher for newer consumer articles where the off-
gassing period dominates the exposure versus older consumer articles that have already undergone the
off-gassing period. Sensitivity analyses indicated that the initial SVOC concentration in the article (a
product of the article density and the weight fraction) is a driver of inhalation exposures for insulation.
For more information on the inhalation exposure estimates, see Section 5.1.2.2.2.
Lifetime Cancer Risks
Inhalation from wood resin composites (e.g., wood flooring) presents the highest lifetime cancer risk
(3.19xl0~2), followed by inhalation exposure from insulation (2,55/10 2) (Table 5-69). In comparing
inhalation risks among wood resin composites, wood flooring has a larger cancer inhalation risk
estimate by two orders of magnitude than wood TV stands. This suggests that the space (surface area) a
wood article occupies in the home environment has a relationship to the magnitude of inhalation risk.
Lifetime cancers risks for wood resin composites (e.g., wood flooring) is dominated by the inhalation
route whereas lifetime cancer risks for wood resin composites (e.g., wooden TV stands) is dominated by
the ingestion of dust. This may be explained by the relatively large surface area for wood flooring vs.
wooden TV stands. Wood articles (e.g., wood flooring, wooden TV stands) have a higher lifetime cancer
risk for dermal exposures (6,35x ]0 3 and 2,03/10 3) compared to oral exposure (4,85/10 4 and
3.99xl0~4). Carpet and foam products (e.g., mattresses, foam furniture, automobile foams) are
dominated by oral cancer risks relative to other routes. The contribution of mouthing exposure from
these articles at younger life stages may be contributing to the overall cancer risk.
Page 331 of 638
-------
Table 5-68. Acute and Chronic Non-cancer Consumer Risk Summary
cou
Consumer
Use Scenario
Exposure
Route
Age Group
(years)
Non-cancer MOEsfl
Overall Confidence
Non-cancer MOEs
Life Cycle
Stage/Category
Subcategory
Acute MOE
UFs = 30
Chronic MOE
UFs = 30
Consumer use/
Furnishing,
cleaning,
treatment, and
care products
Fabric and textile
products
Carpet back
coating
Oral
Infant: 1-2
35
10
Moderate
Oral
Infant: <1
18
5
Textile for
children's
outdoor play
structures
Inhalation
Adult: >21
9
45
Moderate
Oral
Infant: 1-2
30
10
Oral
Infant: <1
17
5
Foam seating and
bedding products
Foam auto
Oral
Infant: 1-2
35
10
Moderate
Oral
Infant: <1
18
5
Foam living
room
Oral
Infant: 1-2
35
10
Slight
Oral
Infant: <1
18
5
Mattress
Oral
Infant: 1-2
35
10
Slight
Oral
Infant: <1
18
5
Foam-other (toy
block)
Oral
Child: 6-10
88
26
Slight
Oral
Child: 3-5
52
15
Oral
Infant: 1-2
7
2
Oral
Infant: <1
4
1
Consumer use/
Construction,
paints, electrical,
and metal
products
Building/
construction
materials -
Insulation
Roofing
insulation
Inhalation
Adult: >21
<1
2
Slight
Oral
Child: 6-10
21
75
Oral
Child: 3-5
7
27
Oral
Infant: 1-2
8
30
Oral
Infant: <1
10
37
Acoustic ceiling
Inhalation
Adult: >21
2
24
Page 332 of 638
-------
cou
Consumer
Use Scenario
Exposure
Route
Age Group
(years)
Non-cancer MOEsfl
Overall Confidence
Non-cancer MOEs
Life Cycle
Stage/Category
Subcategory
Acute MOE
UFs = 30
Chronic MOE
UFs = 30
Consumer use/
Construction,
paints, electrical,
and metal
products
Building/
construction
materials - Wood
and engineered
wood products -
Wood resin
composites
Wood flooring
Dermal
Adult: >1
18
11
Slight
Dermal
Youth: 16-20
19
12
Dermal
Youth: 11-15
17
11
Dermal
Child: 6-10
14
9
Dermal
Child: 3-5
11
7
Dermal
Infant: 1-2
9
6
Dermal
Infant: <1
8
5
Inhalation
Adult: >21
<1
3
Oral
Child: 6-10
17
95
Oral
Child: 3-5
6
48
Oral
Infant: 1-2
6
9
Oral
Infant: <1
6
5
Wooden TV
stand
Dermal
Child: 6-10
93
27
Moderate
Dermal
Child: 3-5
75
22
Dermal
Infant: 1-2
65
19
Dermal
Infant: <1
55
16
Inhalation
Adult: >21
8
416
Oral
Infant: 1-2
34
10
Oral
Infant: <1
18
5
11 Risk estimates are only presented for COUs, routes, and age groups that are below the non-cancer risk benchmarks or above the lifetime cancer benchmarks.
Page 333 of 638
-------
Table 5-69. Lifetime Cancer Consumer Risk Summary
cou
Consumer Use Scenario
Exposure Route
Lifetime Cancer Risk
Estimates"
Overall Confidence in
Cancer Risk Estimate
Life Cycle
Stage/Category
Subcategory
Consumer use/
Furnishing, cleaning,
treatment, and care
products
Fabric and textile
products
Carpet back coating
Dermal
9.24E-06
Moderate
Inhalation
1.48E-04
Oral
3.99E-04
Foam seating and
bedding products
Foam automobile
Dermal
1.52E-04
Moderate
Inhalation
2.52E-08
Oral
3.99E-04
Foam living room
Dermal
3.38E-04
Moderate
Inhalation
4.51E-08
Oral
3.99E-04
Mattress
Dermal
4.12E-05
Slight
Inhalation
2.15E-06
Oral
3.99E-04
Consumer use/
Construction, paints,
electrical, and metal
products
Building/construction
materials - Insulation
Roofing insulation
Dermal
1.85E-05
Slight
Inhalation
2.55E-02
Oral
1.29E-04
Acoustic ceiling
Dermal
1.94E-06
Slight
Inhalation
3.63E-03
Oral
1.43E-05
Building/construction
materials - Wood and
engineered wood
products - Wood
resin composites
Wood flooring
Dermal
6.35E-03
Slight
Inhalation
3.19E-02
Oral
4.85E-04
Wooden TV stand
Dermal
2.03E-03
Moderate
Inhalation
2.08E-04
Oral
3.99E-04
11 Risk estimates are only presented for COUs, routes, and age groups that are below the non-cancer risk benchmarks or above the lifetime cancer benchmarks.
Page 334 of 638
-------
5.3.2.2.2 COUs Without Quantitative Risk Estimates
Paints and Coatings Including Those Found on Automotive Articles and Replacement Parts
Domestic retail production and manufacturing of paints and coatings, including those found on
automotive articles and replacement parts containing TCEP, has ceased, and consumers can no longer
purchase these products from store shelves in the United States. However, TCEP containing paints and
coatings are still used in commercial applications, and consumers could potentially buy commercial
paints and coatings containing TCEP on-line. In addition, consumers may have old cannisters of paints
and coatings that were purchased prior to the phase out of TCEP in the consumer market. Consumers are
unlikely to obtain TCEP containing paints and coatings because the domestic retail production and
manufacturing of TCEP containing paints and coatings has ceased and TCEP containing paints and
coatings represent a small fraction of the paints and coatings on the market.
In the early 2000s, Ingerowski et al. (2001) detected TCEP in 85 percent of 983 household products in
Germany and reported TCEP in wood preservation coatings at 1.0 percent. Also, Haumann and
Thumulla (2002) detected TCEP in paints at a maximum of 840 mg/kg (0.084%) in Germany prior to
2002 (TERA. 2013). Commercial paints and coatings containing TCEP have weight fractions up to 25
percent (Section 3.3.1 of the Engineering Supplemental File Appendix C.4.13), whereas
communications with the Alliance for Automotive Innovation indicated that TCEP is being used in
smaller weight fractions ranging from 0.1 percent to maximum 1 percent.
If a consumer obtains a TCEP-containing paint or coating, exposure could occur through inhalation and
dermal exposure from the application (e.g., spray applied or using a paint brush) of the paint or coating;
or inhalation, oral and dermal exposure from the dried paint or coating through dust generated from
abrasion of the painted or coated article.
Application Scenario
There is a dearth of information regarding the consumer application of paints and coatings exposure
scenario. No consumer SDS are available for TCEP in consumer paints and coatings. EPA has
determined that it is not likely for consumers to obtain TCEP containing paints and coatings products
that are available for commercial applications. Therefore, EPA does not expect exposure to consumers
from the application of TCEP containing paints and coatings.
Dried Paint or Coating Scenario
The dried paint or coating scenario is akin to the article scenarios (e.g., wood resin articles) already
modeled in the TCEP risk evaluation (Section 5.3.2.1.1). EPA is uncertain whether the TCEP used in
paints and coatings is bonded to the polymer matrix (covalently bonded), or whether the TCEP is used
as an additive (non-covalently bonded). For the 2-part resin OES, EPA expects TCEP releases and
dermal exposures to be limited by TCEP being entrained into the hardened polymer matrix (covalently
bonded). If TCEP is entrained in a hardened polymer matrix, a chronic exposure is more relevant than
an acute scenario as the article would either need to be abraded, degrade or the TCEP would have to
migrate from the entrained polymer to the surface of the article and subsequently volatilize. An acute
exposure scenario is more applicable when TCEP is used as an additive, where the TCEP could quickly
off-gas. How TCEP is bonded in the paint or coating, or in its subsequent application may influence the
likelihood of TCEP being released from the consumer product or article.
CEM 3.2 does not include input parameters to specify whether TCEP is bonded to the polymer matrix or
used as an additive. Rather, CEM 3.2 is a set of deterministic models that use physical chemical
properties to estimate inhalation, dermal and ingestion exposures to consumer products and articles. The
Page 335 of 638
-------
CEM results in Figures Figure 5-20, Figure 5-21 and Figure 5-22, describe how inhalation exposures of
articles containing TCEP are expected to be initially high and then subsequently off gas. Older articles
in the home may have already undergone off-gassing of TCEP; thus, there is uncertainty as to the
relevance of continued inhalation exposure from older consumer articles containing TCEP as much of
the exposure may have already occurred in the first few weeks to months.
Consumers use automotive replacement parts that have been painted with TCEP containing paints and
coatings well after the physical painting and coating has occurred in industrial and commercial
applications. However, the chronic exposure scenario (abrasion or delayed release from entrained
polymer matrix) may be more relevant, as TCEP is likely used for its intumescent properties in
automotive applications. Consumers with long commutes, truck drivers, and others who remain in the
indoor vehicle cabin for longer durations may have greater potential for exposure than consumers with
shorter exposure durations in vehicle cabins.
Due to limited reasonably available information regarding the use of articles containing dried paints and
coatings, including those found on automotive articles and replacement parts, and the uncertainties
surrounding the weight fraction, activity, and use patterns, and duration of use for consumers, EPA did
not quantitatively assess the consumer use of articles containing dried paints and coatings including
those used in automotive articles and replacement parts.
To qualitatively evaluate articles containing dried paints and coatings for consumers, EPA looked at the
consumer analysis for other articles (e.g., wood resin articles) containing TCEP. EPA's consumer
analysis for articles containing TCEP, resulted in no chronic inhalation, ingestion, or dermal risk for
adults for the COUs with moderate confidence. However, the consumer analysis did reveal chronic
dermal risk for wood resin composites, and ingestion risk for multiple articles for infants and children.
Based on the similarity of the exposure scenarios, EPA expects articles containing dried paints and
coatings, including those found on automotive articles and replacement parts containing TCEP, to mirror
the other consumer use articles (e.g., wood resin articles) scenarios assessed in Section 5.3.2.2.1. EPA
expects dermal and ingestion risk to infants and children from the use of articles containing dried paints
and coatings, including those found on automotive articles and replacement parts containing TCEP, due
to its similarity to the other consumer article scenarios (e.g., wood resin articles) with an overall
confidence of slight.
Disposal of Wastewater, Liquid Wastes, and Solid Wastes
Consumers may be exposed to articles containing TCEP during disposal and the handling of waste. The
removal of articles in DIY scenarios may lead to direct contact with articles and the dust generated from
the articles. EPA believes that the monitoring data found for the commercial COU of e-waste recycling
would represent similar exposures that could occur during the removal and/or disposal of other articles
containing TCEP. Risk to workers was not found during these activities and therefore it is not expected
that risk would be found in a DIY scenario involving the removal and/or disposal of TCEP containing
articles.
5.3.2.3 Summary of Risk Estimates for the General Population
5.3.2.3.1 COUs with Quantitative Risk Estimates
EPA quantitatively assessed human exposures to TCEP concentrations via oral ingestion of drinking
water, soil, and fish, dermal exposures to soil and surface water, and inhalation of ambient air. EPA
assessed risk associated with each of these exposure scenarios by comparing doses to acute, short-term,
Page 336 of 638
-------
and chronic human equivalent concentrations and doses. Furthermore, EPA assessed the lifetime cancer
risk from TCEP exposure via these routes. As noted previously, EPA uses a range of cancer benchmarks
from 1 in 10,000 to 1 in 1,000,000 to characterize lifetime cancer risks for the general population.
Table 5-70 and Table 5-71 summarize the MOEs used to characterize acute non-cancer risks for oral
exposures for the applicable COUs. Table 5-72 and Table 5-73 summarizes the chronic non-cancer
MOE estimates for the applicable COUs. Table 5-74 summarizes the lifetime cancer oral risk for the
applicable COUs. Oral ingestion non-cancer MOEs and cancer risks are presented for drinking water,
diluted drinking water, landfill leachate to groundwater and subsequent migration to drinking water,
incidental ingestion during swimming, fish ingestion, and soil ingestion for children playing with soil.
Table 5-75 summarizes the acute and chronic non-cancer dermal MOEs for incidental dermal exposures
during swimming and dermal ingestion of soils for children playing with soil associated with applicable
COUs.
Table 5-76 presents the general population chronic inhalation MOEs used to characterize risk for the
applicable COUs. Table 5-77 presents the general population lifetime cancer inhalation risk estimates
for the applicable COUs. Inhalation MOEs and risk estimates are provided for various distances from a
hypothetical facility for two meteorology conditions (Sioux Falls, South Dakota, for central tendency
meteorology; and Lake Charles, Louisiana, for higher-end meteorology).
Ingestion
Drinking Water and Incidental Surface Water Ingestion: Table 5-70 summarizes the acute drinking
water risk estimates for all COUs and lifestages. The non-cancer MOE values for the acute drinking
water ingestion exposure by infants for four scenarios—Incorporation into paints and coatings (1-part
coatings), Incorporation into paints and coatings (2-part coatings), Use in paints and coatings at job sites,
and Formulation of TCEP containing reactive resin—are less than the benchmark MOE of 30. When
factoring in dilution, none of the lifestages have acute drinking water MOE of less than the benchmark
for any scenario.
Because TCEP is recalcitrant to drinking water treatment removal processes, a 0 percent drinking water
treatment removal efficiency was used to calculate the oral drinking water exposure doses. The non-
diluted acute risk estimates assume the general population was drinking water at the site of the facility
outfall. To approximate a more typical drinking water concentration, distances between drinking water
intake locations and facilities based on SIC codes were used to calculate a dilution factor to estimate a
diluted drinking water concentration (see Section 5.1.3.4.1). All non-cancer MOEs from acute incidental
ingestion via swimming were larger than the benchmark MOE of 30 for adults, youth, and children (see
Appendix I).
None of the chronic MOEs from drinking water, diluted drinking water, incidental ingestion via
swimming, and drinking water contamination from landfill leachate were lower than the benchmark
MOE of 30. Drinking water MOEs are presented for both diluted and non-diluted surface water
concentrations. The diluted drinking water MOEs represent typical case scenarios, whereas MOEs based
on the non-diluted concentrations represent worst-case scenarios.
The DRAS model described in Section 3.3.3.8 estimated TCEP groundwater concentrations from
landfill leachate. Only two industrial and commercial release scenarios have anticipated releases to
landfill (Incorporation into paints and coatings - 1-part coatings and Processing into 2-part resin article).
The DRAS model estimated groundwater concentrations by a range of loading rates and a range of
leachate concentrations rather than the release estimate generated by the two industrial uses (21.5
Page 337 of 638
-------
kg/site-year for 1-part coatings, and 42.9 kg/site-year for 2-part resin article). Nevertheless, estimates via
the full production volume did not result in chronic oral MOEs below 30 for drinking water.
Lifetime (from birth) oral ingestion cancer risk greater than 1 in 1,000,000 is associated with releases
from four OESs: (1) Incorporation into paints and coatings - 1-part coatings; (2) Incorporation into
paints and coatings - Resins/sol vent-borne; (3) Use in paints and coatings at job sites; and (4)
Processing into 2-part resin article. There was also oral ingestion cancer risk greater than 1 in 1,000,000
for the adult lifetime for the same scenarios, except for the use in paints and coatings at job sites. Under
diluted drinking water conditions, no lifetime risks from birth or for the adult lifetimes exceeded 1 in
1,000,000.
Fish Ingestion: For the adult general population, acute exposure estimates via fish ingestion using a
BAF of 2,198 L/kg showed MOEs less than 30 for all OESs except laboratory use of chemicals (Table
5-32). No OESs had an acute risk estimate less than 30 based on a BAF of 109 L/kg. For the adult
subsistence fisher, EPA only had one fish IR that resulted in the same doses for both acute and chronic
exposure. EPA estimated non-cancer MOEs by comparing that same dose with both the acute and
chronic HEDs. Exposure estimates based on a BAF of 2,198 L/kg showed MOEs less than the acute
benchmark for all OESs except laboratory use of chemicals. Using a BAF of 109, Laboratory use of
chemicals and import and repackaging showed MOEs above the acute benchmark. For Tribes, the same
approach was to estimate acute and chronic risks as the subsistence fisher. A BAF of 2,198 showed
MOEs less than the acute benchmark for all OESs except Laboratory use of chemicals based on the
current IR and for all OESs based on the heritage IR. A BAF of 109 showed MOEs less than the acute
benchmark for all COUs except Import and repackaging and Laboratory use of chemicals based on the
current mean IR (for the Suquamish Tribe). The BAF of 109 also had MOEs less than the acute
benchmark for all COUs except Laboratory use of chemicals based on the heritage IR (for the Kootenai
Tribe).
Chronic exposure for the general population resulted in MOEs less than the chronic benchmark of 30 for
all OESs except Laboratory use of chemicals for both fish IRs and a BAF of 2,198/kg (Table 5-72). The
table presents adult general population risk estimates based on only the 90th percentile IR even though
two values were used, as discussed in Section 5.1.3.4.2. The MOEs based on the central tendency IR
will be 4.4 times higher. When estimating exposure and risks based on a BAF of 109 L/kg, there are
some differences in risks between the two IRs. The 90th percentile IR results in risks below their chronic
benchmark for three OESs: (1) Incorporation into paints and coatings - 1-part coating; (2) Incorporation
into paints and coatings - 2-part reactive coatings; and (3) Formulation of TCEP containing reactive
resin. The central tendency IR did not result in any OESs with risk estimates below their chronic
benchmark.
Chronic exposure for the subsistence fisher and Tribes resulted in MOEs less than 30 for all OESs based
on a BAF of 2,198 L/kg and all IRs. A BAF of 109 L/kg showed risk estimates less than the chronic
benchmark for all OESs except Laboratory use of chemicals.
Exposure estimates were not calculated for younger age groups. For younger age groups, acute and
chronic MOEs less than benchmark values are reasonably expected because these age groups generally
have higher fish ingestion rates per kilogram body weight (Table Apx 1-2). For Tribes, adults were
reported to have the highest IR per kilogram of body weight (see Section 2265.1.3.4.4).
For the adult general population, subsistence fisher, and Tribe, cancer risk estimates are above 1 in
1,000,000 for all OESs and for both BAF values, as well as current and heritage IRs for Tribes. Table
Page 338 of 638
-------
5-75 shows the lifetime cancer risk estimates for fish ingestion. Cancer risk estimates were not
calculated for fish ingestion among younger age groups. Similar to non-cancer risk, cancer risks for
younger age groups are reasonably expected to be higher than older groups because of the higher fish
ingestion rate per kilogram of body weight or because adults have the highest IR by body weight.
(TableApx 1-2).
The above risk estimates are based on the harmonic mean surface water flows representing the 50th
percentile stream flows of all facilities in each industry sector. EPA also estimated exposure and risks
using the 90th percentile stream flow of the harmonic mean. Acute and chronic non-cancer risk
estimates are above their corresponding benchmarks for most of the COUs for all population groups
using the 90th percentile stream flow. However, cancer risks estimates exceed 1 in 1,000,000 for almost
all the COUs, population groups, and two BAFs. Results are presented in Table 5-75. These results
indicate the critical role of receiving water flow as an input in determining TCEP concentrations in
surface water and thus exposure via fish ingestion.
Soil Ingestion: Chronic oral non-cancer MOEs from soil were estimated for children 3 to 6 years of age
based on soil concentrations that were calculated from air deposition for various distances from a
hypothetical facility releasing TCEP (see Section 3.3.3.2). Oral doses were calculated for two exposure
scenarios: (1) a child conducting activities with soil, and (2) a child playing in mud (see Section
5.1.3.4.4). No MOEs were less than the benchmark of 30 for the children's soil ingestion scenario for
any of the COUs. In addition, there was no lifetime cancer risk for soil ingestion for any of the COUs.
Dermal
Incidental Dermal from Swimming: Non-cancer MOEs were not lower than benchmark values for the
acute and chronic incidental dermal exposures swimming scenario for any of the COUs.
Children's Dermal Exposure from Playing in Soil: Dermal exposure estimates from soil were estimated
for children 3 to 6 years of age because these ages are expected to play in mud and perform activities
with soil. Soil concentrations were calculated via annual air deposition fluxes for various distances from
a hypothetical facility releasing TCEP (see Section 3.3.3.2). Dermal exposure doses were also calculated
for a child conducting activities with soil and a child playing in mud (see Section 5.1.3.3.2). No non-
cancer MOEs for chronic exposures were less than the benchmark MOE of 30 at 100 or 1,000 m for
either scenario of children playing in mud or children conducting activities with soil.
Many uncertainties are associated with the dermal exposure estimate used for the chronic dermal MOE
that was less than the benchmark, including the lack of release information, site information, and
reasonableness of the exposure scenario. The source of the exposure is a hypothetical facility that
releases TCEP to the air for 2 days. Because no site information was available, EPA's release
assessment estimated a 50th percentile of 27 sites to a 95th percentile of 203 sites per the OES for the
commercial use of paints and coatings. To observe an MOE less than the benchmark, a child would have
to be playing in mud at 100 m from the hypothetical facility. TCEP would deposit to the soil after
deposition from air releases. Section 3.3.3.2 describes how EPA calculates soil concentrations from
annual modeled air deposition. No U.S. studies recorded TCEP in soil. Modeled soil concentrations at
100 m (4.15 x 103 ng/g) were two orders of magnitude higher than the TCEP concentrations found in
Germany (23.5 ng/g) (Mihailovic and Fries. 2012). The study from Germany also indicated increased
soil concentration of TCEP due to snow melt (see Section 3.3.3.1).
Page 339 of 638
-------
Inhalation
Table 5-78 shows the COUs where EPA found lifetime inhalation cancer risk estimates greater than 1 in
1,000,000 for the 2,500 lb production volume, high-end release estimate, suburban forest scenario and
when using both central-tendency and high-end meteorological data. EPA found inhalation cancer risks
greater than the benchmark for the 50th percentile air concentrations for the use of paints and coatings at
job sites at distances as far as 100 m from the site. EPA also found cancer risk above this benchmark for
the 95th percentile air concentrations for the use of paints and coatings out to 100 m from the job site.
Table 5-77 displays the chronic inhalation non-cancer risk estimates for the 2,500 lb production volume,
high-end release estimate, suburban forest scenario, high-end meteorological data at 10 m from the
facility. No non-cancer inhalation MOEs were less than the acute (total UF = 30) or chronic (total UF =
30) benchmark MOEs for any COUs. The lowest MOE for the chronic exposure scenario was 498 (the
use of paints and coatings scenario, high meteorological station data, at 10 m, 95th percentile). The
lowest MOE for the acute exposure scenario was 295,000 for the Processing into 2-part resin article,
high meteorological station data, at 10 m, 95th percentile scenario (not shown). Ambient air is a minor
environmental compartment as described in Section 2.2.
It is unlikely that individual residences will be within 10 m of the stack or fugitive air release from these
facilities. However, these estimates suggest that fence line communities living within 100 m downwind
of facilities that use TCEP in paints and coatings at job sites may be at an increased risk of developing
cancer over their lifetimes.
When compared to the monitoring literature, the maximum modeled ambient air concentrations 2.55
ng/m3, are within an order of magnitude of the ambient concentrations described in Bradman et al.
(2014) that recorded a maximum concentration of 1.60 ng/m3, mean of 0.72 ng/m3, at 14 early childhood
education facilities in California between May 2010 and May 2011. For deposition, Moran et al. (2023)
reports 1.36><10_6 g/m2/day which equates to 4.74xl0~4 g/m2 a year. These modeled deposition values
are one to two orders of magnitude higher than the high-end deposition values observed 1,000 m from
the hypothetical releasing facility for the Use of Paints and Coatings - Spray Application OES, 2,500 lb
Production Volume, 95th Percentile Release Estimate, Suburban Forest Land Category Scenario
(9.51xl0~3 to 1.47><10~2 g/m2 a year).
Page 340 of 638
-------
Table 5-70. General Population Acute Drinking Water (Oral Ingestion) Non-cancer Risk Summary
cou
OES
Acute Oral Non-cancer MOEs
UFs = 30
Drinking Water
Drinking Water (Diluted)
Life Cycle
Stage/Category
Subcategory
Adult
(>21 yr)
Infant
(21 yr)
Infant
(
-------
Table 5-71. Acute Fish Ingestion Non-cancer Risk Summary Based on 50th Percentile Flow of Harmonic Mean
cou
OES
Acute Oral Non-cancer MOEs
UFs = 30
General Population
Subsistence Fishers
Tribes
(Current IR)"
Tribes
(Heritage IR)*
Life Cycle Stage/
Category
Subcategory
BAF
2,198
BAF 109
BAF
2,198
BAF
109
BAF
2,198
BAF
109
BAF
2,198
BAF
109
Manufacturing/
Import
Import
Repackaging
18
363
3
57
2
37
<1
5
Processing/Processing
-Incorporation into
formulation, mixture,
or reaction product
Flame retardant
in: paint and
coating
manufacturing
Incorporation into
paints and coatings - 1-
part coatings
4
82
1
13
<1
8
<1
1
Incorporation into
paints and coatings - 2-
part reactive coatings
4
90
1
14
<1
9
<1
1
Polymers used
in aerospace
equipment and
products
Formulation of TCEP-
containing reactive
resin
3
65
<1
10
<1
7
<1
1
Commercial
use/Laboratory
chemicals
Laboratory
chemical
Use of laboratory
chemicals
450
9,067
70
1,411
46
930
6
122
Commercial and
Industrial use/Paints
and coatings
Paints and
coatings
Use of paints and
coatings at job sites
8
154
1
24
1
16
<1
2
" Current fish consumption rate at 216 g/day based on survey of Suquamish Indian Tribe in Washington (see Section 5.1.3.4.4).
b Heritage fish consumption rate at 1,646 g/day based on study of Kootenai Tribe in Idaho (see Section 5.1.3.4.4).
Page 342 of 638
-------
Table 5-72. General Populai
ion Chronic Water and Soil Ingestion Non-cancer Risk Summary
cou
Chronic Non-cancer Oral MOEs
UFs = 30
Life Cycle
Stage/C atego ry
Subcategory
OES
Drinking
Water
(Diluted)
Drinking
Water
Drinking
Water (via
Leaching to
Groundwater)
Ambient
Water
(Incidental
Ingestion)
Soil Intake
(50th)
at 100 m
Soil Intake
(95th)
at 100 m
Soil Intake
(50th)
at 1,000 m
Soil Intake
(95th)
at 1,000 m
Manufacturing/
Import
Import
Repackaging
1.64E08
1.05E05
N/A
2.11E05
2.20E10
5.15E09
1.73E12
4.03E11
Incorporation into
paints and
4.40E07
23,728
2.12E06
4.89E04
7.02E08
1.64E08
7.95E10
1.86E10
Flame retardant
coatings - 1 -part
Processing/
in: paint and
coatings
Processing -
Incorporation
coating
manufacturing
Incorporation into
paints and
4.85E07
26,171
N/A
5.39E04
4.85E09
1.13E09
3.68E11
8.59E10
into formulation,
coatings - 2-part
mixture, or
reactive coatings
reaction product
Polymers used in
aerospace
equipment and
products
Formulation of
TCEP containing
reactive resin
9.89E06
18,706
N/A
4.62E04
4.41E09
1.03E09
3.46E11
8.07E10
Processing/
Processing -
Incorporation
into article
Aerospace
equipment and
products and
automotive
articles and
replacement parts
containing TCEP
Processing into 2-
part resin article
N/A
N/A
2.12E06
N/A
5.15E08
1.20E08
5.05E10
1.18E10
Commercial
use//Laboratory
chemicals
Laboratory
chemical
Use of laboratory
chemicals
4.10E09
2.60E06
N/A
5.30E06
4.60E08
1.07E08
4.20E10
9.81E09
Commercial and
Industrial
Paints and
Use of paints and
coatings at job
6.96E07
4.47E04
N/A
8.98E04
2.98E05
6.96E04
5.72E07
1.34E07
use/Paints and
coatings
sites
coatings
Page 343 of 638
-------
Table 5-73. Chronic Fish Ingestion Non-cancer Risk Summary
cou
General Population
Subsistence Fishers4
Tribes (Current)'
Tribes (Heritage)''
Life Cycle
Subcategory
OES
BAF 2,198a
BAF 109a
BAF 2,198
BAF
BAF
BAF
BAF
BAF
Stage/C atego ry
CTe
HE
CTe
HE
109
2,198
109
2,198
109
Manufacturing/
Import
Import
Repackaging
23
5
461
105
1
16
1
11
<1
1
Incorporation into
5
1
104
24
<1
4
<1
2
<1
<1
Flame
paints and coatings
Processing/Proc
essing -
Incorporation
into formulation,
mixture, or
reaction product
retardant in:
- 1 -part coatings
paint and
coating
manufacturing
Incorporation into
paints and coatings
- 2-part reactive
coatings
6
1
115
26
<1
4
<1
3
<1
<1
Polymers used
Formulation of
4
1
82
19
<1
3
<1
2
<1
<1
m aerospace
equipment and
products
TCEP containing
reactive resin
Aerospace
Processing into 2-
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Processing/Proc
equipment and
products and
part resin article
essing -
Incorporation
into article
automotive
articles and
replacement
parts containing
TCEP
Commercial
Laboratory
Use of laboratory
571
130
11,505
2,617
20
407
13
268
2
35
use/Laboratory
chemicals
chemical
chemicals
Commercial and
Paints and
Use of paints and
10
2
196
45
<1
7
<1
5
<1
1
Industrial
use/Paints and
coatings
coatings at job sites
coatings
" General population exposure estimates based on general population fish ingestion rate of 22.2 g/day.
4 Subsistence fishers exposure estimates based on subsistence fisher ingestion rate of 142.2 g/day.
c Current fish consumption rate at 216 g/day based on survey of Suquamish Indian Tribe in Washington (see Section 5.1.3.4.4).
'' Heritage fish consumption rate at 1,646 g/day based on study of Kootenai Tribe in Idaho (see Section 5.1.3.4.4).
* Exposure estimates based on a general population mean fish ingestion rate of 5.04 g/day.
Page 344 of 638
-------
Table 5-74. General Population Lifetime Cancer Oral Ingestion Risk Summary Table
cou
OES
Lifetime Cancer Oral Risk Estimates
Drinking Water
Drinking Water
(Diluted)
Life Cycle Stage/Category
Subcategory
Lifetime
from Birth
Adult
Lifetime
Lifetime
from Birth
Adult
Lifetime
Manufacturing/Import
Import
Repackaging
6.09E-07
6.37E-07
3.91E-10
4.09E-10
Processing/Processing - Incorporation
into formulation, mixture, or reaction
product
Flame retardant in: paint
and coating manufacturing
Incorporation into paints and coatings -
1-part coatings
2.70E-06
2.82E-06
1.45E-09
1.52E-09
Incorporation into paints and coatings -
2-part reactive coatings
2.44E-06
2.56E-06
1.32E-09
1.38E-09
Processing/Processing-Incorporation
into formulation, mixture, or reaction
product
Polymers used in
aerospace equipment and
products
Formulation of TCEP containing
reactive resin
1.43E-06
1.50E-06
9.20E-10
9.63E-10
Commercial use/Laboratory chemicals
Laboratory chemical
Use of laboratory chemicals
5.44E-08
5.69E-08
3.49E-11
3.65E-11
Commercial and Industrial use/Paints
and coatings
Paints and coatings
Use of paints and coatings at job sites
3.42E-06
3.58E-06
6.47E-09
6.77E-09
Page 345 of 638
-------
Table 5-75. Lifetime Cancer Risk Summary for Fish Consumption
cou
OES
Lifetime Cancer Oral Risk Estimates
Life Cycle
Stage/
Category
Subcategory
Adult Fish Ingestion General Population"
Adult Subsistence
Fisher
Tribes
(Current IR)
Tribes
(Heritage IR)
BAF 2,198
BAF 109
BAF
2,198
BAF
109
BAF
2,198
BAF
109
BAF
2,198
BAF
109
CTA
HE
CT"
HE
Manufacturing/
Import
Import
Repackaging
2.32E-03
1.02E-02
1.15E-04
5.07E-04
6.57E-02
3.26E-03
9.96E-02
4.94E-03
7.59E-01
3.77E-02
Processing/
Processing -
Incorporation
into formulation,
mixture, or
reaction product
Flame retardant
in: paint and
coating
manufacturing
Incorporation into
paints and coatings
- 1 -part coatings
1.03E-02
4.53E-02
5.11E-04
2.25E-03
2.91E-01
1.44E-02
4.41E-01
2.19E-02
3.36
1.67E-01
Incorporation into
paints and coatings
- 2-part reactive
coatings
9.34E-03
4.11E-02
4.63E-04
2.04E-03
2.64E-01
1.31E-02
4.00E-01
1.98E-02
3.05
1.51 E—01
Polymers used in
aerospace
equipment and
products
Formulation of
TCEP containing
reactive resin
1.31E-02
5.74E-02
6.48E-04
2.85E-03
3.69E-01
1.83E-02
5.60E-01
2.78E-02
4.27
2.12E-01
Commercial
use/Laboratory
chemicals
Laboratory
chemical
Use of laboratory
chemicals
9.32E-05
4.10E-04
4.62E-06
2.03E-05
2.63E-03
1.31E-04
3.99E-03
1.98E-04
3.04E-02
1.51E-03
Commercial and
Industrial
use/Paints and
coatings
Paints and
coatings
Use of paints and
coatings at job sites
5.47E-03
2.41E-02
2.71E-04
1.19E-03
1.55E-01
7.67E-03
2.35E-01
1.16E-02
1.79
8.87E-02
" Cancer risk estimates for the adult general population are based on the high-end fish ingestion rate of 22.2 g/day.
4 Exposure estimates are based on a general population mean fish ingestion rate of 5.04 g/day.
Page 346 of 638
-------
Table 5-76. General Population Dermal Acute and Chronic Non-cancer Risk Summary
cou
OES
Acute MOEs
UFs = 30
Chronic Non-cancer MOEa
UFs =30
Life Cycle
Stage/Category
Subcategory
Surface Water
(Adult
Swimming)
Surface Water
(Adult
Swimming)
Child
Playing in
Mud at
100 ma
Child
Activities with
Soil at 100 ma
Child Playing
in Mud at
1,000 ma
Child Activities
with Soil at
1,000 ma
Manufacturing/Import
Import
Repackaging
6.82E03
4.55E05
6.95E06
1.43E09
5.44E08
1.12E11
Processing/Processing
-Incorporation into
formulation, mixture,
or reaction product
Flame retardant
in: paint and
coating
manufacturing
Incorporation into paints
and coatings - 1 -part
coatings
1.54E03
1.05E05
2.21E05
4.55E07
2.51E07
5.15E09
Incorporation into paints
and coatings - 2-part
reactive coatings
1.70E03
1.14E05
1.53E06
3.14E08
1.16E08
2.39E10
Polymers used in
aerospace
equipment and
products
Formulation of TCEP
containing reactive resin
1.21E03
9.75E04
1.39E06
2.86E08
1.09E08
2.24E10
Processing/Processing
-Incorporation into
article
Aerospace
equipment and
products and
automotive
articles and
replacement parts
containing TCEP
Processing into 2-part
resin article
N/A
N/A
1.62E05
3.34E07
1.59E07
3.27E09
Commercial
use/Laboratory
chemicals
Laboratory
chemical
Use of laboratory
chemicals
1.70E05
1.13E07
1.45E05
2.98E07
1.33E07
2.72E09
Commercial and
Industrial use/Paints
and coatings
Paints and
coatings
Use of paints and coatings
at job sites
2.90E03
1.95E05
9.4E01
1.93E04
1.80E04
3.71E06
" Soil concentration based of annual air deposition fluxes is used to estimate the acute exposures scenario of a child playing with mud and conducting activities in soil.
Page 347 of 638
-------
Table 5-77. Inhalation Chronic Risk Summary for General Population"
cou
OES
Chronic Inhalation MOEs
UFs = 30
Life Cycle Stage/Category
Subcategory
Ambient Air 50th
Ambient Air 95th
Manufacturing/Import
Import
Repackaging
9.34E07
5.10E07
Processing/Processing -
Incorporation into formulation,
mixture, or reaction product
Flame retardant in:
paint and coating manufacturing
Incorporation into paints and coatings -
1-part coatings
3.66E06
1.49E06
Incorporation into paints and coatings -
2-part reactive coatings
2.22E07
7.18E06
Polymers used in aerospace
equipment and products
Formulation of TCEP containing reactive
resin
1.98E07
6.41E06
Processing/Processing -
Incorporation into article
Aerospace equipment and products
and automotive articles and
replacement parts containing TCEP
Processing into 2-part resin article
2.41E06
1.82E06
Commercial use/Laboratory
chemicals
Laboratory chemical
Use of laboratory chemicals
2.10E06
1.48E06
Commercial and Industrial
use/Paints and coatings
Paints and coatings
Use of paints and coatings at job sites
1.23E03
4.98E02
" 2,500 lb Production Volume - High-End Release Estimate, Suburban Forest Scenario at 10 m, high meteorology conditions
Page 348 of 638
-------
Table 5-78. General Population Lifetime Cancer Inhalation Risk Summary Table"
cou
OES
Distances
(m)
Lifetime Cancer Inhalation Risk
Life Cycle
Stage/Category
Subcategory
Central Tendency
Meteorological Data
High-End Meteorological Data
Cancer Risk
Estimate for 50th
Percentile Air
Concentration
Cancer Risk
Estimate for 95th
Percentile Air
Concentration
Cancer Risk
Estimate for 50th
Percentile Air
Concentration
Cancer Risk
Estimate for 95th
Percentile Air
Concentration
Commercial
and Industrial
use/Paints and
coatings
Paints and
coatings
Use in paints
and coatings
at job sites
10
4.55E-05
5.53E-05
5.42E-05
1.33E-04
30
1.49E-05
2.50E-05
1.50E-05
3.74E-05
30-60
6.74E-06
1.70E-05
7.16E-06
2.29E-05
60
4.63E-06
1.07E-05
4.76E-06
1.20E-05
100
1.66E-06
4.66E-06
1.69E-06
4.46E-06
" 2,500 lb Production Volume - High-End Release Estimate, Suburban Forest Scenario
Page 349 of 638
-------
5.3.2.3.2 COUs Without Quantitative Risk Estimates
Distribution in Commerce
Distribution in commerce includes transporting TCEP or TCEP-containing products between work sites
or to final use sites, as well as loading and unloading from transport vehicles. The general population
may be in the proximity of vehicles that transport TCEP or TCEP-containing products.
TCEP production volumes have declined and recent reports (e.g., the 2020 CDR) indicate that
production volumes may be below reporting levels; therefore, the precise volume is unknown. The
general decline in production volume would logically lead to decreased distribution into commerce.
Therefore, exposure and risk would also likely have declined with time. Exposure is possible from
ongoing manufacturing, processing, industrial, and commercial uses. However, EPA lacks the data to
assess the full set of risks to the general population from this COU. Due to this limited data, EPA's
confidence in these exposures is indeterminate. Nevertheless, given that TCEP and/or TCEP containing
products or articles are expected to be transported in sealed containers or packages; EPA anticipates that
exposure and releases during Distribution in Commerce will be negligible.
Processing — Recycling
EPA did not quantify risks to the general population from releases during recycling of either e-waste or
recycled foam products due to limited reasonably available information and limited use of TCEP in
electronics.
EPA did not find data to quantify releases of TCEP from e-waste recycling facilities. The total releases
are expected to be low for several reasons: The volume of TCEP in e-waste products is low; only a
fraction of the products is recycled; and recycling will likely be dispersed over many e-waste sites.
Although EPA located information on the presence of TCEP at e-waste recycling facilities during
systematic review, the data sources did not provide the volume of TCEP-contained electronics processed
at any of the facilities identified. Therefore, EPA's confidence in these exposures is indeterminant and
cannot quantify risk from e-waste recycling. However, EPA acknowledges that while some releases and
exposures could occur during the disposal of the wide variety of items that TCEP has found its way into,
these are expected to be minimal and dispersed, and exposures are expected to be negligible.
TCEP may be present within flexible foam, fabric, textile, and other applications that have been made
from recycled foam scraps generated during trimming of original TCEP-containing manufactured foam
products. EPA was not able to determine, with reasonable accuracy, the exact flame retardants that are
used in these products and did not locate information on releases during recycling of such foam.
Industrial and Commercial Use (Other) — Aerospace Equipment and Products and Automotive
Articles and Replacement Parts Containing TCEP
EPA does not expect significant releases to the environment for the following COUs:
• Industrial use - Other use - Aerospace equipment and products and automotive articles and
replacement parts containing TCEP; OES: installing article (containing 2-part resin); and
• Commercial use - Other use - Aerospace equipment and products and automotive articles and
replacement parts containing TCEP; OES: installing article (containing 2-part resin).
After TCEP-containing resins have cured within products that are installed, EPA expects TCEP releases
and dermal exposures will be limited by TCEP being entrained into the hardened polymer matrix.
During installation it is possible that very small levels of dust could be generated, these were quantified
in Table 5-67. and do not indicate risk to workers from inhalation nor do they indicate the generation of
Page 350 of 638
-------
significant dust releases occurring. Releases may occur via the mechanism of blooming (volatilization
from the cured resin surface) during the service life of the article, but EPA expects that such releases
during installation will be negligible (OECD. 2009; NICNAS. 2001). Installation of aerospace
equipment and products would be installed without any type of further processing of the article that
would lead to potential releases (sanding, drilling, etc.). Therefore, the potential risk to workers and the
general population from releases during installation of TCEP-containing articles is low.
Commercial Uses that TCEP Is No Longer Actively Incorporated Into
EPA determined that the following commercial use COUs for TCEP consists of use of existing products
and end of service life disposal:
• Commercial use - Furnishing, cleaning, treatment/care products - Fabric and textile products;
• Commercial use - Furnishing, cleaning, treatment/care products - Foam seating and bedding
products;
• Commercial use - Construction, paint, electrical, and metal products - building/construction
materials - Insulation; and
• Commercial use - Construction, paint, electrical, and metal products - Building/construction
materials - Wood and engineered wood products - Wood resin composites.
These COUs were being phased out beginning in the late 1980s or early 1990s and replaced by other
flame retardants or flame-retardant formulations. EPA did not locate data to estimate (1) the amount of
TCEP that was historically used in these products, (2) the amounts of these products that have already
reached the end of their service life, or (3) the amounts of these products that have already been
disposed. Based on the years that the phase-out occurred, many of these products not likely to be in use
because the end of their service life was already reached (e.g., commercial roofing has an estimated
lifespan of 17-20 years). EPA assumes that any of these products still used commercially represent a
fraction of the overall amount of TCEP previously used for these purposes. Therefore, releases to the
environment from these commercial uses would also represent only a fraction of previous release
amounts.
The consumer assessment for these articles resulted in no consumer risk for inhalation, ingestion dermal
risk for adults for the COUs with moderate confidence. Therefore, under similar exposure durations and
exposure frequencies, EPA expects these articles to pose no commercial risk for inhalation, ingestion,
and dermal risk to COUs to the general population who use articles in a similar fashion to consumers.
Furthermore, exposure to the releases from the disposals of these articles are anticipated to be less than
the consumer use of these articles in the home.
Disposal
Disposal is possible throughout the life cycle of TCEP and TCEP-containing products, including waste
treatment and disposal resulting from manufacturing, processing, commercial and consumer uses.
For processing COUs, EPA estimated releases to landfills or incinerators (see Section 5.3.2.1) for
• Incorporation into formulation, mixture, or reaction product - Paint/coating manufacture - 1-part
coating OES (landfill);
• Incorporation into articles - Aerospace equipment and products and automotive articles and
replacement parts containing TCEP - Processing in two-part resin article OES (landfill);
• Incorporation into formulation, mixture, or reaction product - Paint/coating manufacture - 2-part
coating OES (incineration); and
Page 351 of 638
-------
• Incorporation into formulation, mixture, or reaction product - Polymers in aerospace equipment
and products - Formulation of reactive resins OES (incineration).
Both releases to landfills and incinerators rely on inputs provided by ESDs or GSs, but the ESDs and
GSs do not specify the proportion of the throughput that goes to either of these two disposal practices.
Therefore, EPA was unable to further quantify environmental releases related to these two disposal
processes. For three of these processing COUs, EPA was able to perform quantitative risk
characterization for releases to surface water (which includes on-site wastewater treatment or discharge
to POTWs, where applicable) (see Table 3-2); any releases to on-site waste treatment or POTWs were
combined with other exposures and this combined risk to the general population was quantified for these
processing COUs.
Waste treatment (POTW or on-site) or disposal (landfill or incineration) is expected to be negligible for
industrial and commercial uses related to installing articles. For the COUs of
Manufacturing/repackaging, Commercial use of paints and coatings, and Commercial use of laboratory
chemicals, Disposal to landfills or incinerators is not expected but EPA estimated surface water releases
that could include release to wastewater treatment or POTWs and any resulting risks to the general
population were assessed for the individual COUs.
For the commercial uses that have been phased out, any currently used products that contain TCEP are
expected to be disposed in landfills but will represent just a fraction of previous amounts when TCEP
was used more widely. Landfills would likely contain TCEP in commercial articles from these COUs,
but data are lacking with which to estimate exposure and risk from disposal or waste treatment activities
for these COUs, and EPA has not quantified such risks. For e-waste recycling, there is also too little
information to estimate exposure from disposal and only a small portion of e-waste is expected to
contain TCEP.
The DRAS model described in Section 3.3.3.8 estimated TCEP groundwater concentrations from
landfill leachate. Only two industrial and commercial release scenarios have anticipated releases to
landfill (Incorporation into paints and coatings - 1-part coatings and Processing into 2-part resin article).
The DRAS model estimated groundwater concentrations by a range of loading rates and a range of
leachate concentrations rather than the release estimate generated by the two industrial uses (21.5
kg/site-year for 1-part coatings, and 42.9 kg/site-year for 2-part resin article). Nevertheless, estimates via
the full production volume did not result in chronic oral MOEs below 30 for drinking water.
While this analysis estimates potential exposures to the general population via disposal to landfills, this
approach does not capture all the types of disposals that may occur throughout the manufacturing,
processing, distribution, import, industrial use, commercial use and consumer use life cycle of TCEP
containing products and articles. There may be releases to the environment from consumer articles
containing TCEP via end-of-life disposal and demolition of consumer articles in the built environment,
and the associated down-the-drain release of TCEP from domestic laundry that removes TCEP
containing dust from clothing to wastewater. It is difficult for EPA to quantify these end-of-life and
down-the-drain laundry exposures due to limited reasonably available information on source attribution
of the consumer COUs.
EPA's confidence in these general population exposures from the disposal scenario is indeterminant.
The Agency acknowledges that while some releases and exposures could occur during the disposal of
the wide variety of items that TCEP has been incorporated into, these exposures are expected to be
negligible.
Page 352 of 638
-------
5.3.2.4 Summary of Risk Estimates for Infants from Human Milk
EPA estimated infant risks from milk ingestion based on TCEP concentrations in milk modeled for
maternal exposures associated with consumer, occupational, and general population groups. Infant
exposures through milk were estimated for both mean (105 mL/kg-day) and upper (153 mL/kg-day)
milk intake rates. Risk estimates for intermediate and chronic infant exposures through milk were
calculated for both cancer and non-cancer endpoints for each COU within each maternal group. Short-
term risks, which have an averaging time of 30 days or less, were estimated based on the infant's first
month of life. The first month of life generally had the highest doses because of the highest milk
ingestion rate per kilogram of body weight; thus, it is most protective for estimating intermediate risks.
For chronic non-cancer risks, exposure typically occurs over at least 10 percent of lifetime in adults.
However, it cannot be ruled out that continuous exposure during the first year of life will result in
permanent health effects through adulthood. Chronic risks were thus considered for infant doses in the
first year of life. Similarly, cancer risks were also estimated using the linear low-dose extrapolation even
though exposure did not occur over the lifetime.
Acute infant doses were not estimated because the Verner Model is designed to estimate TCEP
concentrations in milk and doses from continuous exposure rather than an acute, 1-day dose. However,
if short-term or chronic doses result in risk estimates below their corresponding benchmark MOEs, EPA
estimated acute risks by comparing short-term and chronic doses with an acute POD. Appendix 1.5.1
through Appendix 1.5.5 presents risk estimates for all iterations that EPA considered.
For the consumer exposure pathways, intermediate and chronic infant risk estimates were above the
corresponding benchmark MOEs for all COUs. Infant cancer risk estimates are above 1 in 1,000,000 for
two consumer exposure scenarios regardless of milk intake rate: (1) Building/construction materials not
covered elsewhere (roofing insulation), and (2) Building/construction materials - Wood and engineered
wood products (wood flooring). The infant cancer risk estimates for these two COUs range from
7.04xl0~6 to 1.03xl0~5. The maternal cancer risk estimates for the same COUs range from 8.11xl0~6 to
4.5 x 10~2 (Table 5-69). Although the lower bound of the cancer risk estimates for the mother and infant
are similar, it is important to note that maternal risks are calculated by separate exposure routes (i.e.,
oral, dermal, and inhalation). Dermal exposure to roofing insulation resulted in the lowest maternal
cancer risk estimates, and all other routes resulted in risk estimates that were two to four magnitudes
higher. Other COUs with cancer risk estimates above 1 in 1,000,000 for the mother were below this
level for the infant ingesting human milk. Therefore, infant risks are not proportionally higher than
maternal risks. Furthermore, the maternal risk estimates in Table 5-69 are based on doses for an adult
weighing 80 kg. If they were adjusted for women of reproductive age, the risk estimates for this
population will increase given the higher dose. This underscores the conclusion that minimizing
maternal exposure to TCEP is most important for protecting an infant, as the mother is more sensitive.
For the occupational exposure pathways, 1- and 2-day application of spray paints and coatings were not
evaluated because the Verner model is intended to estimate only continuous maternal exposure. Among
the evaluated OESs, short-term and chronic infant risk estimates were below their benchmark MOEs for
Commercial use - Paints and coatings - Spray (2-part coatings, 250-day application) regardless of the
maternal dose type (chronic or subchronic) and milk intake rate (mean or upper). For Laboratory
chemicals, a mean milk intake rate resulted in short-term risk estimates below their benchmark MOEs
based on a subchronic maternal dose. An upper milk intake rate for the same OES resulted in short-term
and chronic infant risk estimates below their benchmark MOEs regardless of the maternal dose type.
Lastly, for Incorporation into paints and coatings - 1-part coatings, a mean milk intake rate resulted in
short-term risk estimates below their benchmark MOEs based on a subchronic maternal dose. An upper
milk intake rate and subchronic maternal dose for the same OES resulted in short-term and chronic
Page 353 of 638
-------
infant risk estimates below the benchmark MOE. However, acute infant risk estimates were above the
MOE for all of the above OESs.
Cancer risk estimates vary depending on the maternal worker dose type and the milk intake rate. For
subchronic maternal doses, infant cancer risk estimates exceeded 1 in 1,000,000 for 8 out of the 10
OESs regardless of milk intake rate:
• Import and repackaging;
• Incorporation into paints and coatings - 1-part coatings;
• Incorporation into paints and coatings - 2-part reactive coatings;
• Processing - Formulation of TCEP into 2-part reactive resins;
• Processing - Processing into 2-part resin article;
• Commercial use - Paints and coatings - Spray (1-part, 250-day application);
• Commercial use - Paints and coatings - Spray (2-part reactive coatings, 250-day application);
and
• Laboratory chemicals.
For the above OESs, infant cancer risk estimates ranged from 2.67x 10~6 to 6.06x 10~5. The OES that
showed short-term and chronic infant risks also showed the highest infant cancer risk estimates:
commercial use - paints and coatings - spray (2-part coatings, 250-day application). For this OES,
infant cancer risk estimates based on a mean and upper milk intake rate were 3.61 x 10~5 and 6.06x 10~5,
respectively.
For chronic maternal doses, infant cancer risk estimates exceeded 1 in 1,000,000 for five or seven OESs,
depending on the milk intake rate:
• Import and repackaging (only for upper milk intake rate);
• Incorporation into paints and coatings - 1-part coatings;
• Processing - Formulation of TCEP into 2-part reactive resins (only for upper milk intake rate);
• Processing - Processing into 2-part resin article;
• Commercial use - Paints and coatings - Spray (1-part coatings, 250-day application);
• Commercial use - Paints and coatings - Spray (2-part reactive coatings, 250-day application);
and
• Laboratory chemicals.
For the above OESs, infant cancer risk estimates ranged from 1,06x 10~6 to 4.91 x 10~5. Again,
Commercial use - Paints and coatings - Spray (2-part coatings, 250-day application) had the highest
infant cancer risk estimate at 3.37x 10~5 and 4.91 x 10-5 for a mean and upper milk intake rate,
respectively. Overall, for occupational exposure pathways, the risk estimates for short-term, chronic, and
cancer effects are lower in the infants compared to the mothers.
EPA estimated risks to infants in Tribal communities where the mothers were exposed to TCEP through
fish ingestion. As discussed in Section 5.1.3.4.4, a current mean ingestion rate (IR) and heritage IR was
used. The milk intake rate (mean vs. upper) did not significantly change risk estimates. For the high
BAF, both milk intake rates and both fish IRs resulted in MOEs below the short-term and chronic
benchmarks for all OESs except Laboratory use of chemicals. All OESs had cancer risk estimates above
1 in 1,000,000. The low BAF and current IR did not show any MOEs below the short-term and chronic
benchmarks for all OESs and both milk intake rates. However, cancer risks exceeded 1 in 1,000,000 for
all OESs except Laboratory use of chemicals. The low BAF, heritage IR, and mean milk intake rate
resulted in risk estimates below the short-term and chronic benchmarks for the same three OESs, plus
Page 354 of 638
-------
one OES with only short-term risk estimates below the benchmark. Cancer risks exceeded 1 in
1,000,000 for the low BAF, heritage IR, and mean milk intake rate for all OESs except Laboratory use
of chemicals. Lastly, the OESs that had MOEs below the short-term and chronic benchmarks were also
compared against the acute benchmark to determine if there are acute risks at that exposure level. A high
BAF had MOEs below the acute benchmark (3-5 OESs depending on the fish and milk IR type). A low
BAF had no risk estimates below the acute benchmark except for one scenario.
For the general population, EPA focused on maternal oral exposures because they resulted in
significantly higher doses than dermal or inhalation. Within the oral routes, ingestion of fish (at the
general population's 90th percentile IR of 22.2 g/day) and undiluted drinking water were among the
sentinel pathways for mothers. EPA estimated infant risks using these pathways and did not combine
across other routes. Using a low BAF, no OESs had short-term or chronic risk estimates below the MOE
based on the mean and upper milk uptake rate. Cancer risk estimates did not exceed 1 in 1,000,000 for
any of the OESs based on the mean intake rate. However, based on the upper milk intake rate, the cancer
risk estimate for Formulation of TCEP containing reactive resin did exceed 1 in 1,000,000 (1.21 xlCT6).
For the general population adult fish ingestion based on the high BAF, no OESs had risk estimates
below their short-term and chronic MOEs for both milk intake rates. Cancer risk estimates exceeded 1 in
1,000,000 for all OESs except Laboratory use of chemicals. Under the mean milk intake rate, cancer risk
estimates ranged from 2.96x 10~6 to 1,66x 10~5. Under the upper milk intake rate, cancer risk estimates
ranged from 4.32><10~6 to 2.43xl0~5. The OES with the highest cancer risk estimate is Formulation of
TCEP containing reactive resin. Risk estimates for infants of subsistence fisher were not calculated but
are expected to fall in between those for the adult general population and Tribal population.
Due to the uncertainties in estimating fish ingestion exposure as discussed in Section 5.3.2.3, EPA also
considered ingestion of undiluted drinking water. This pathway did not result in any non-cancer risk
estimates below the benchmark MOE or cancer risk estimates above 1 in 1,000,000. No maternal risks
were observed either. While it is possible that combining other exposure routes, such as dermal
absorption from swimming, can result in additional scenarios showing infant risk estimates below their
benchmark MOEs, results from consumer, occupational, and general population fish ingestion
demonstrated that the mothers are more sensitive than the infants. There are no COUs or OESs across all
maternal groups that showed higher risk estimates in the infants compared to the mothers. In fact, some
COUs resulted in maternal doses and risk estimates that are several magnitudes higher for the mothers
than the infants. Therefore, protecting the mother will also protect the infant from exposure via human
milk.
5.3.3 Risk Characterization for PESS
EPA considered PESS throughout the exposure assessment and throughout the hazard identification and
dose-response analysis. The Agency has identified several PESS factors that may contribute to a group
having increased exposure or biological susceptibility. Examples of these factors include lifestage,
occupational and certain consumer exposures, nutrition, and lifestyle activities.
For the TCEP risk evaluation, EPA accounted for the following PESS groups: infants exposed through
human milk from exposed individuals, children and male adolescents who use consumer articles or are
among the exposed general population, subsistence fishers, Tribal populations, pregnant women,
workers and consumers who experience aggregated or sentinel exposures, people who live in fenceline
communities near facilities that emit TCEP, and firefighters.
Page 355 of 638
-------
The non-cancer hazard values used in the risk evaluation are based on pregnant rats that may be more
susceptible to neurotoxic effects and male reproductive endpoints.
Table 5-79 summarizes how PESS were incorporated into the risk evaluation and also summarizes the
remaining sources of uncertainty related to consideration of PESS. Appendix D provides additional
details on PESS considerations for the TCEP risk evaluation.
Page 356 of 638
-------
Table 5-79. Summary of PESS Considerations Incorporated into the Risk Evaluation
PESS
Categories
Potentially Exposed Individuals
Susceptible Subpopulations
Potential Increased Exposures
Incorporated into Exposure
Assessment
Sources of Uncertainty for
Exposure Assessment
Potential Sources of Biological
Susceptibility Incorporated into
Hazard Assessment
Sources of Uncertainty for
Hazard Assessment
Lifestage
• Lifestage-specific exposure
scenarios included infants
exposed through human
milk.
• Exposure factors by age
group were applied to
calculate consumer oral and
dermal exposures.
• Children scenarios of playing
in mud and activities with
soil considered for dermal
and oral soil ingestion.
• Mouthing of consumer
articles considered for infants
and children.
• The level of exposure via
milk is uncertain as described
in Section 5.1.3.7.2.
• Uncertainties regarding the
appropriateness for adjusting
inhalation values to younger
lifestages for the consumer
analysis.
• There is potential susceptibility related
to different lifestages using adolescent
male mice as the POD for intermediate
and chronic exposure. Potential
differences in other lifestages, such as
older individuals, which might relate
toxicokinetic or toxicodynamic
differences was addressed through a
10x UF for human variability (see
Section 5.2.7 for POD and UFs).
• The intermediate/chronic POD is
expected to be protective of adolescent,
developmental, and adult outcomes
(including pregnant females) based on
comparison with existing
developmental and reproductive
studies and a 2-year bioassay for
TCEP. Pregnant females are the basis
of the acute POD.
• Human data lend slight evidence of
possible developmental effects in
infants and children.
• The magnitude of
differences in toxicokinetics
and toxicodynamics for
some individuals may be
greater than accounted for
by the UFh of 10.
• Inability to use some
reproductive/developmental
data due to errors in one
study results in uncertainty
regarding the magnitude of
some effects in offspring.
• Some uncertainty exists
based on limited number of
studies and differences in
specific outcomes among
studies.
Pre-existing
Disease
• EPA did not identify pre-
existing disease factors
influencing exposure.
• Pre-existing diseases and conditions,
especially those that lead to
neurological and behavioral effects,
reproductive effects, and cancer may
increase susceptibility to the effects of
TCEP.
• This greater susceptibility is addressed
through the 10x UF for human
variability.
• The increase in
susceptibility is not known
and is a source of
uncertainty; differences may
be greater than the UFh of
10.
Page 357 of 638
-------
PESS
Categories
Potentially Exposed Individuals
Susceptible Subpopulations
Potential Increased Exposures
Incorporated into Exposure
Assessment
Sources of Uncertainty for
Exposure Assessment
Potential Sources of Biological
Susceptibility Incorporated into
Hazard Assessment
Sources of Uncertainty for
Hazard Assessment
Lifestyle
Activities
• EPA evaluated exposures
resulting from subsistence
fishing and considered
increased intake of fish in
these populations, as well as
Tribal populations.
• There is a high level of
uncertainty in the BAF
values because of limited
monitoring data. There is also
uncertainty in the modeled
surface water concentrations.
• EPA did not identify lifestyle factors
that specifically influence
susceptibility to TCEP and that could
be quantified. Generally, certain
factors (e.g., smoking, alcohol
consumption, diet) can affect health
outcomes.
• This is a remaining source
of uncertainty.
Occupational
and
Consumer
Exposures
• Monitoring data suggest that
firefighters have elevated
TCEP exposures because of
firefighting activities
(indicated by elevated urine
concentrations of BCEP, a
metabolite of TCEP (Maver et
al.. 2021; Javatilaka et al..
2017)).
• Consumer articles intended for
use by children (children's
play structures, toy foam
blocks) considered in the
assessment of COUs.
• Uncertainties in duration of
use of consumer articles in
the home.
• EPA did not identify occupational and
consumer exposures that influence
susceptibility.
• This is a remaining source
of uncertainty.
Socio-
demographic
• EPA did not evaluate exposure
differences between racial
groups.
• Monitoring literature
indicates TCEP levels in dust
are significantly associated
with the presence of
extremely worn carpets. This
may be relevant for lower
socioeconomic status
families (Castorina et al..
2017).
• Slight evidence that sociodemographic
factors may influence susceptibility to
TCEP (decreased IQ).
• There is still some
uncertainty given the
limited, slight evidence.
Nutrition
• EPA did not identify
nutritional factors influencing
exposure.
• Nutrition can affect susceptibility to
disease generally. EPA did not identify
specific evidence that nutritional
factors influence susceptibility to
TCEP.
• This is a remaining source
of uncertainty.
Page 358 of 638
-------
PESS
Categories
Potentially Exposed Individuals
Susceptible Subpopulations
Potential Increased Exposures
Incorporated into Exposure
Assessment
Sources of Uncertainty for
Exposure Assessment
Potential Sources of Biological
Susceptibility Incorporated into
Hazard Assessment
Sources of Uncertainty for
Hazard Assessment
Genetics/
Epigenetics
• EPA did not identify genetic
or epigenetic factors
influencing exposure.
• Genetic disorders may increase
susceptibility to male reproductive
effects; this was addressed through a
10x UF for human variability (see
Section 5.2.5.1.2).
• The magnitude of the impact
of genetic disorders is
unknown and is a source of
uncertainty; differences may
be greater than the UFh of
10.
Unique
Activities
• EPA did not evaluate activities
that are unique to Tribal
populations (e.g., sweat
lodges, powwows). The
evaluation of high fish
consumption among Tribal
populations is included in the
category Lifestyle Activities.
• There is uncertainty in how
exposure factors (e.g., water
consumption rate) change for
specific Tribal lifeways.
• EPA did not identify unique activities
that influence susceptibility.
• This is a remaining source
of uncertainty.
Page 359 of 638
-------
Potentially Exposed Individuals
Susceptible Subpopulations
PESS
Categories
Potential Increased Exposures
Incorporated into Exposure
Assessment
Sources of Uncertainty for
Exposure Assessment
Potential Sources of Biological
Susceptibility Incorporated into
Hazard Assessment
Sources of Uncertainty for
Hazard Assessment
Aggregate
Exposures
• Occupational dermal and
inhalation exposures
aggregated.
• Consumer inhalation, dermal,
and oral ingestion exposures
are presented by individual but
are aggregated in Appendix J.
• Uncertainty is associated
with several exposures that
EPA did not aggregate (see
Section 5.1.4):
o Inhalation and drinking
water for the general
population from co-
located facilities due to the
lack of reasonably
available site-specific
data for TCEP.
o Across consumer,
commercial, or industrial
COUs due to a lack of data
indicating such co-
exposures exist for TCEP.
Across exposure scenarios
based on release estimates
for the general population
because such assumptions
could result in double-
counting. Across other
exposure scenarios (e.g.,
mouthing consumer
articles, drinking water)
due to a lack of data
indicating the co-exposure
of TCEP.
• Not relevant to susceptibility
Other
Chemical
and Non-
chemical
Stressors
• EPA did not identify factors
influencing exposure.
• In vitro data on co-exposure with
benzo[a]pyrene showed increased
impacts on inflammation and
proliferation pathways.
• TCEP showed anti-estrogenic activity
in vitro after co-exposure with 17(3-
estradiol.
• There is insufficient data to
quantitatively address
potential increased
susceptibility due to these
factors; this is a remaining
source of uncertainty.
Page 360 of 638
-------
EPA considered susceptibility when conducting hazard identification and dose-response analysis for
TCEP. Some observations may be made regarding factors that may increase susceptibility to the effects
of TCEP. Human data suggest there may be susceptible subpopulations, although as identified in Section
5.2.3, human evidence is only slight or indeterminate. Percy et al. (2022) found increased BCEP in urine
was associated with lower IQ in children with SES using more than one measure related to SES.
Developmental effects related to growth and gestational age show male infants alone and both sexes of
offspring were less likely to be small for their gestational age in one study (Oh et al.. 2024) and have
increased skinfold thickness (two measures of thickness for males; one measure for both sexes) in
another study (Crawford et al.. 2020). Female children had a greater incidence of being pre-term (Oh et
al.. 2024) and lower birthweight and length (Yang et al.. 2022). Effects may differ by gender, as
identified by some epidemiological studies, including the developmental effects on growth and
gestational age and possible greater susceptibility by female children to hay fever/allergies (Mendv et
al.. 2024).
Animal studies showed slight evidence for developmental effects (NTP. 1991a). EPA identified some
sensitive sexes for certain health outcomes (higher incidence of neurotoxicity in female rats (NTP.
1991b). greater sensitivity of male (vs. female) mice in reproductive effects (Chen et al.. 2015a)). and
EPA quantified risks based on male reproductive effects in the risk evaluation. An acute POD based on
neurotoxicity was identified for pregnant rats (Moser et al.. 2015).
As identified in TableApx E-2, other susceptibility factors are generally considered to increase
susceptibility of individuals to chemical hazards. These factors include pre-existing diseases, alcohol
use, diet, stress, among others. The effect of these factors on susceptibility to health effects of TCEP is
not known; therefore, EPA is uncertain about the magnitude of any possible increased risk from effects
associated with TCEP exposure.
For non-cancer endpoints, EPA used a default value of 10 for human variability (UFh) to account for
increased susceptibility when quantifying risks from exposure to TCEP. The Risk Assessment Forum, in
A Review of the Reference Dose and Reference Concentration Processes (U.S. EPA. 2002b). discusses
some of the evidence for choosing the default factor of 10 when data are lacking and describe the types
of populations that may be more susceptible, including different lifestages (e.g., of children and elderly).
U.S. EPA (2002b). however, did not discuss all the factors presented in Table Apx E-2. Thus,
uncertainty remains regarding whether these additional susceptibility factors would be covered by the
default UFh value of 10 chosen for use in the TCEP risk evaluation.
For cancer, the dose-response model applied to animal tumor data employed low-dose linear
extrapolation, and this assumes any TCEP exposure is associated with some positive risk of getting
cancer. EPA made this assumption in the absence of an established MOA for TCEP and according to
guidance from U.S. EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005b). Assuming
all TCEP exposure is associated with some risk is likely to be health conservative because EPA does not
believe that a mutagenic MOA is likely for TCEP. However, information is lacking with which to
determine whether there is an MOA that acts via a non-linear dose-response. Even though the cancer
dose-response modeling assumes any exposure is associated with a certain risk, EPA presents risk
estimates in comparison with benchmark risk levels (1 in 1,000,000 to 1 in 10,000).
Although there is likely to be variability in susceptibility across the human population, EPA did not
identify specific human groups that are expected to be more susceptible to cancer following TCEP
exposure. Other than relying on animal tumor data for the more sensitive sex, the available evidence
does not allow EPA to evaluate or quantify the potential for increased cancer risk in specific
Page 361 of 638
-------
subpopulations, such as for individuals with pre-existing diseases or those who smoke cigarettes. Given
that a mutagenic mode of action is unlikely, EPA does not anticipate greater cancer risks from early life
exposure to TCEP. Therefore, EPA is not applying an age-dependent adjustment factor.
EPA also considered PESS throughout the exposure assessment. EPA estimated infant risks from milk
ingestion based on TCEP concentrations in milk modeled for maternal exposures associated with
consumer, occupational, and general population groups. Infant exposures through milk were estimated
for both mean (105 mL/kg-day) and upper (153 mL/kg-day) milk intake rates. Risk estimates for
intermediate and chronic infant exposures through milk were calculated for both cancer and non-cancer
endpoints for each COU within each maternal group. Although EPA only had slight confidence in the
exposure estimates for infants for this pathway, EPA did determine that infants exposed through human
milk ingestion are not more sensitive than the mothers. Protecting the mother will also protect the infant
from exposure via human milk. Results of that analysis are included in Section 5.3.2.4.
For the general population, EPA also identified subsistence fishers, children, infants, and people who
live in fenceline communities as PESS groups. In its evaluation, EPA considered the increased intake of
fish in subsistence fishers. Although there was not enough reasonably available information to assess
exposures for Tribal populations specifically, EPA quantitatively evaluated the Tribal fish ingestion
pathway for TCEP. Children, infants, and people who live in fenceline communities were also identified
as a PESS group for the general population through the drinking water pathway and soil ingestion
pathways. The fish ingestion analysis and the analysis of children's exposure through drinking water and
soil can be found in Section 5.3.2.3.1.
For occupational exposures, EPA also conducted a qualitative assessment for firefighters. Monitoring
data suggests that firefighters have elevated TCEP exposures as a result of firefighting activities.
Elevated levels of flame retardants have been found in dust collected from fire stations and in firefighter
personal equipment (Shen et al.. 2018). A study on firefighters reported increased urine concentrations
of BCEP, a metabolite of TCEP, from pre-fire to 3- and 6-hour post fire collections. Although the results
were not statistically significant, pre-fire vs. post fire concentrations indicate that firefighters may be at
increased risk of TCEP exposures during structure fires (Mayer et al.. 2021). Researchers from the CDC
measured urine samples for BCEP in 76 members of the general population and 146 firefighters who
performed structure firefighting while wearing full protective clothing and SCBA respirators. BCEP was
detected in 10 percent of the general population at a median level that was below the detection limit and
in 90 percent of firefighters at a median of 0.86 ng/mL (Javatilaka et al.. 2017). TCEP was measured at
five fire stations across the United States (California, Minnesota, New Hampshire, New York, and
Texas) at median concentrations of 1,040 ng/g. In comparing chemical concentrations by vacuum use,
this study did not observe any differences in TCEP concentrations due to cleaning practices (vacuuming)
(Shen et al.. 2018). These levels are less than the median (2,700 ng/g) concentrations measured in 2011
in California house dust (Dodson et al.. 2012). The U.S. Fire Profile study states that the total number of
firefighters in 2020, 364,300 (35%) were career, while 676,900 (65%) were volunteers. The U.S. Fire
Profile study also states that the number of fire departments for career firefighters is up to a total of
5,244 establishments and a total of 24,208 establishments for volunteer firefighters (NFPA. 2022).
For consumer exposures, EPA identified and evaluated the exposure for PESS groups including children
and infants through exposure to consumer products. Risk estimates for these PESS groups can be found
in Section 5.3.2.2. EPA has moderate confidence in the fabric and textile products COU, and slight to
moderate confidence in the foam seating and bedding products and building/construction materials-
wood resin COUs. Confidence ratings are derived from consideration of variety of factors including
Page 362 of 638
-------
confidence in the model used, the default values, and the input parameters (e.g., density, use duration,
weight fraction, dermal parameters), and the corroborating monitoring data (see Table 5-18).
Limited information was available on the TCEP COUs. However, the Ecology Washington database
(WSDE. 2023) sampled consumer articles that children under 3 years of age are expected to contact
and/or mouthed. Of the 268 products related to TSCA COUs, 24 articles were detected to have TCEP.
Eleven out of twenty-four (4 percent of total) articles were related to fabric and textiles uses, whereas 13
out of 24 (5 percent of total) were in foam articles. Products were sampled in the summer of 2012.
Ionas et al. (2014) sampled children's toys in Antwerp, Belgium, and reported an overall detection
frequency of 28 percent (32 out of 114) of TCEP detected in children toys produced around the year
2007. Two out of eight articles were for wooden toys. Fang et al. (2013) reported a detection frequency
of 95 percent (19 out of 20) of V6/TCEP in vehicles with an average model year of 2004. Stapleton et al.
(2012) detected only one instance of V6/TCEP in 102 foam couches across the United States during
2011-2012.
Table 5-80. Summary of Detection Frequencies and Sampling Dates for Relevant Consumer
Products Containing TCEP
cou
Detection
Frequency
n
Source
Sampling Date
Life Cycle Stage/
Category
Subcategory
Consumer Use/
Furnishing,
cleaning,
treatment/care
products
Fabric and textile
products
4%
268
Ecology Washington
database (WSDE.
2023)
2012
Foam seating and
bedding products
(foam couches)
1%
102
Stanleton et al. (2012)
2011-2012
5%
268
Ecology Washington
database (WSDE.
2023)
2012
70%
20
Fans et al. (2013)
2009-2011
Foam seating and
bedding products
(auto foam)
95%
20
Fans et al. (2013)
2009-2011
vehicle average
model year 2004
Construction,
paint, electrical,
and metal
products
Building/
construction
materials - Wood
and engineered
wood products -
Wood resin
composites
100%
1
SCHER (2012)
1997
25%
8
Ionas et al. (2014)
2007
Table 5-80 provides a summary of the detection frequencies of the monitoring literature. It is significant
that all these frequency estimates were estimated before the implementation of California TB 117-2013,
and it is anticipated that manufacturers have phased out TCEP from their product due to the introduction
of the less stringent flammability standards for upholstered furniture (TB 117-2013).
Page 363 of 638
-------
Table 5-81. Suggested Consumer Population Sizes Based on Characterization of Consumer Article
Detection Frequencies
COU
Detection
Frequency
Adjusted
Detection
Frequency:
Current
Use
Total U.S.
Population
(of
331,449,281)"
Total U.S.
Children
under 5
years(of
18,400,235)"
Total U.S.
Females of
Reproductive
Age (of
118,273,566)"
Life Cycle
Stage/Category
Subcategory
Furnishing,
cleaning,
treatment/Care
products
Fabric and
textile
products
4%
0.4%
1,325,797
73,601
473,094
Foam seating
and bedding
products
5%
0.5%
1,657,246
92,001
591,368
Construction,
paint, electrical,
and metal
products
Building/
construction
materials -
Wood and
engineered
wood
products -
Wood resin
composites
\%b
1%
3,314,493
184,002
1,182,736
11 Values from the 2020 U.S. Census.
h It was the assessor's judgement to overwrite literature detection frequency value. Only 9 samples presented TCEP
use in wooden products.
Table 5-81 assigns a detection frequency value for each COU above slight-moderate confidence. Four
percent is chosen for Fabric and Textile Products, and five percent is selected for foam seating and
bedding products. Although Fang et al. (2013) indicates higher detection frequencies in vehicles (95%),
the vehicles selected in this study were from an average model year of 2003.5, and it is understood that
auto manufacturers have moved away from using V6/TCEP formulations in their vehicles. A detection
frequency value of one percent is selected for wood resin products, due to the scarce number of
examples indicating TCEP use in wood articles.
An order of magnitude correction to adjust the detection frequencies to current uses was applied for
fabric and textile products and foam seating and bedding products to adjust for TB 117-2013. The
adjustment did not apply to wood resin composites because TB 117-2013 applies to upholstered
furniture.
To characterize the population utilizing these consumer articles, the adjusted detection frequencies are
multiplied by the total U.S. population, total U.S. population of children under 5 years of age, and total
U.S. population of females of reproductive age from the 2020 US census. This calculation provides a
ballpark figure of the expected number of individuals who are exposed to current consumer articles.
Major assumptions in the characterization of this population include the idea that the use of these
consumer articles scale linearly with the detection frequency of detection among consumer articles, the
detection frequencies in the monitoring literature is representative of the use of TCEP compared to other
flame retardants in the marketplace, and that the order of magnitude adjustment is sufficient to reflect
the phase away from TCEP to other OPFRs.
Page 364 of 638
-------
5.3.4 Risk Characterization for Aggregate and Sentinel Exposures
Section 2605(b)(4)(F)(ii) of TSCA requires EPA, as a part of the risk evaluation process, to describe
whether aggregate or sentinel exposures under the COUs were considered and the basis for their
consideration.
The term aggregate is defined as "the combined exposures to an individual from a single chemical
substance across multiple routes and across multiple pathways" in the Agency's final rule, Procedures
for Chemical Risk Evaluation Under the Toxic Substances Control Act (82 FR 33726. July 20, 2017)
(see Appendix A.2).
In the procedural rule, EPA defines sentinel exposure as "the exposure from a chemical substance that
represents the plausible upper bound of exposure relative to all other exposures within a broad category
of similar or related exposures" (40 CFR 702.33). In this evaluation, EPA considered sentinel exposures
by considering risks to populations who may have upper bound exposures, including workers and ONUs
who perform activities with higher exposure potential and people who live in fenceline communities.
EPA characterized high-end exposures using modeling approaches and if available, using monitoring
data. Where information on the distribution of exposures is available, EPA typically uses the 95th
percentile value of the available dataset to characterize high-end exposure for a given COU.
Across Routes
Figure 5-15 aggregates the consumer exposure estimates by route (inhalation, dermal, ingestion) for
each COU and lifestage combination. In addition, this supplemental file includes risk tables that indicate
whether aggregation across routes results in risk. Figure 5-18 and Figure 5-19 provide two examples
where an aggregation across routes could result in chronic and acute risk, whereas consideration from a
single route would not result in risk. For example, for Figure 5-18, if dermal, ingestion, and inhalation
routes were considered individually, the exposure estimates are lower than the intermediate/chronic
HED of 2.73 mg/kg-day or HEC 14.9 mg/m3 divided by the benchmark MOE of 30. However, when
aggregating dermal and inhalation exposures, the toxicity value of 2.73 mg/kg-day divided by the
benchmark MOE is exceeded.
Page 365 of 638
-------
Aggregate Chronic Average Daily Doses (CADDs)
TCEP COUs
0,15
Chronic HED/MOE
Dermal
Ingestion
Inhalation
•v
¦}%
Cn
Oh
'Sfr,
(Jf.
%
COU Lifestage
r-tt
(Jr.
% y
*o.
Or.
%i
X
Figure 5-18. Aggregate CADDs for Consumer Use of Textiles in Outdoor Play Structures at Adult,
Youth2, and Youthl Lifestages
Page 366 of 638
-------
Aggregate Acute Doses (ADRs)
TCEP COUs
Acute HED/MOE
I Dermal
0-& I Ingestion
I Inhalation
5 0.6
textile-carpet Child 1 textile-carpet Infant2
COU Life stage
Figure 5-19. Aggregate ADRs for Carpet Back Coating, Childl, and Infant2 Lifestages
There were no instances of aggregate lifetime risk for any COU where there was not already risk to the
COU from an individual route. The supplemental file includes risk tables that can further be toggled to
explore aggregate risks.
EPA combined exposures for the milk pathway across all routes for each COUs/OESs within workers
and consumers. However, for the general population, EPA only assessed the oral route when assessing
the milk pathway because exposure estimates showed that oral doses were several magnitudes higher
than dermal or inhalation doses. As a result, oral exposures will be the primary driver for infant risks via
the milk pathway. Furthermore, within the adult oral pathways that include fish ingestion, drinking
water ingestion, and incidental water ingestion from swimming, EPA only considered fish and drinking
water ingestion. These two pathways constitute the highest oral doses, thus having the greatest potential
to result in infant risks from human milk ingestion. Indeed, infant cancer risk estimates exceeded 1 in
1,000,000 for all COUs/OESs based on maternal fish ingestion (high BAF). Aggregating other exposure
scenarios will not further inform risk characterization.
Across Exposure Scenario
The confidence in the general population exposure scenarios for drinking water ingestion, fish ingestion
(lowBAF), and inhalation (100 m) is moderate. For the formulation of TCEP containing reactive resin
OES, chronic non-diluted drinking water exposure estimates are 1,46x 10 4 mg/kg/d. For the same OES,
chronic fish ingestion concentrations are two to three orders of magnitude higher for the general
population and subsistence fishers at 0.033 and 0.94 mg/kg/d, respectively. Chronic inhalation exposure
estimates are given in mg/m3 and do not exhibit risk—even at 10 m from a hypothetical facility.
Therefore, aggregate exposure across general population exposure scenarios does not result in an
appreciable difference as the exposure is dominated by the sentinel exposure (fish ingestion).
Page 367 of 638
-------
Furthermore, since the general population and subsistence fisher estimates result in chronic risk for all
COUs, aggregating additional exposure scenarios (e.g., consumer, occupational) with the general
exposure scenarios (fish ingestion) is uninformative in characterizing risks.
The confidence in the consumer COUs is moderate for the subcategories of carpet back coating, textile
in outdoor play structures, living room foam, automobile foam, and wooden TV stands. Chronic
ingestion estimates are above the chronic benchmark (0.091 mg/kg/d) for each of these subcategories
(carpet back coating, textile in outdoor play structures, living room foam, automobile foam, and wooden
TV stands), and chronic dermal estimates are above the benchmark for wooden TV stands. Because the
consumer exposure estimates result in chronic risk, aggregating additional exposure scenarios (e.g.,
general population, occupational) with the consumer exposure scenarios is uninformative in
characterizing risk.
The other consumer exposure scenario subcategories (e.g., insulation, mattress, wood resin) have slight
confidence. Aggregating these subcategories with additional exposure scenarios (e.g., general
population, occupational) would be uninformative in characterizing risk due to the slight confidence in
these scenarios.
5.3.5 Overall Confidence and Remaining Uncertainties in Human Health Risk
Characterization
EPA took fate, exposure (occupational, consumer, and general population), and human health hazard
considerations into account when characterizing the human health risks of TCEP. Human health risk
characterization evaluated confidence from occupational, consumer, and general population exposures
and human health hazards. Hazard confidence and uncertainty is represented by health outcome and
exposure duration as reported in Section 5.2.6, which presents the confidence, uncertainties, and
limitations of the human health hazards for TCEP. Confidence in the exposure assessment has been
synthesized in the respective weight of scientific evidence conclusion sections for occupational
exposures (see Section 5.1.1.4), consumer exposures (see Section 5.1.2.4), and general population
exposures (see Section 5.1.3.7). Table 5-82 provides a summary of confidence for exposures and
hazards for non-cancer endpoints for the COUs that resulted in any non-cancer risks, and Table 5-83
provides a confidence summary for cancer for the COUs that resulted in cancer risks.
Uncertainties associated with the occupational exposure assessment include a lack of reported data from
databases such as TRI, NEI, DMR, and more recently, CDR. Site-specific data were only available for a
small number of current processors, and it is not clear if this data are representative of these industries
and workplace practices.
Uncertainties associated with the general population exposures assessment included the lack of site-
specific information, the incongruence between the modeled concentrations and doses with the
monitoring data, and the complexity of the assessed exposure scenarios. Section 5.1.3.7 illustrates the
confidence in the assessment of the general population exposure scenarios.
5.3.5.1 Occupational Risk Estimates
Exposure Monitoring Data and Use of Models
EPA only identified monitoring data for dust occurring within an e-waste recycling facility. Monitoring
data for the remaining COUs/OESs was not found. Surrogate monitoring data were found to assess
potential exposure to TCEP during installation of articles and this estimated inhalation exposure used
TCEP monitoring data for furniture manufacturing (Makinen et al.. 2009). Surrogate monitoring data are
Page 368 of 638
-------
also used for the assessment of paints and coatings use during spray application. It is unclear if these
COUs have similar worker activities and if they are fully representative of worker exposure for the
OESs of installation of articles and use of paints and coatings. The remaining COUs/OESs used
modelling approaches to estimate worker exposures.
Where sufficient data were reasonably available, the 95th and 50th percentile exposure concentrations
were calculated using these data. The underlying distribution of the data, and the representativeness of
the reasonably available data, are not known. Where discrete data were not reasonably available, EPA
used reported statistics from the Monte Carlo simulations {i.e., 50th and 95th percentile). Because EPA
could not verify these values, there is an added level of uncertainty.
For OESs that do not have monitoring data, EPA used relevant GSs and/or ESDs to identify worker
activities and exposure routes that are reasonably expected to occur. Exposure distributions were then
created using Monte Carlo simulation with 100,000 iterations and the Latin hypercube sampling method.
EPA calculated ADC and LADC values assuming workers and ONUs are regularly exposed during their
entire working lifetime, which likely results in an overestimate. Individuals may change jobs during
their career such that they are no longer exposed to TCEP; therefore, actual ADC and LADC values
would be lower than the estimates presented.
Although EPA has confidence in the models used, it is possible that they may not account for variability
of exact processes and practices at an individual site. Furthermore, there are no 2020 CDR reports for
TCEP and only one from 2016. Therefore, EPA made assumptions about pounds per site-year (2,500 lb
presented in risk tables) that leads to uncertainty in these estimates.
Assumptions Regarding Occupational Non-users
Exposures for ONUs can vary substantially and most data sources do not sufficiently describe the
proximity of these employees to the TCEP exposure source. As such, exposure levels for the
"occupational non-user" category will have high variability depending on the work activity; therefore,
all ONU exposure estimates except for recycling of e-waste are considered to have only slight
confidence. For the OES of recycling of e-waste, monitoring data were available for workers conducting
activities consistent with the activities of ONUs, this results in a confidence rating of moderate to robust.
Modeled Dermal Exposures
The Fractional Absorption Model is used to estimate dermal exposure to TCEP in occupational settings.
The model also assumes a single exposure event per day based on existing framework of the EPA/OPPT
2-Hand Dermal Exposure to Liquids Model and does not address variability in exposure duration and
frequency. Additionally, the studies used to obtain the underlying values of the quantity remaining on
the skin (Qu) did not take into consideration the fact that liquid retention on the skin may vary with
individuals and techniques of application on and removal from the hands. Also, the data used were
developed from three kinds of oils; therefore, the data may not be applicable to other liquids. Based on
these uncertainties, EPA has a moderate level of confidence in the assessed baseline exposure.
Page 369 of 638
-------
Number of Workers
There are several uncertainties surrounding the estimated number of workers potentially exposed to
TCEP. Most are unlikely to result in a systematic underestimate or overestimate but could result in an
inaccurate estimate. CDR data were not available to estimate the number of workers associated with
manufacturing, processing, or use of TCEP. There are also uncertainties with BLS data, which are used
to estimate the number of workers for the remaining COUs. EPA had to use higher-level NAICS codes
(at 3- to 5-digit level) combined with assumptions from the U.S. Census' SUSB, which could result in
inaccuracies if the distribution of workers in occupations with TCEP exposure differs from the overall
distribution of workers in each NAICS. Also, EPA needed to designate which industries and occupations
have potential exposures, and this may result in over- or underestimation. However, any inaccuracies
would not be likely to systematically either overestimate or underestimate the number of exposed
workers.
Weight Fraction Considerations
The COU that had the lowest TCEP concentration weight fraction in the products assessed was the Use
of Paints and Coatings. The weight fraction, from the product SDS's, ranged from 0.1 to 25 percent
(Table 5-14). A Monte Carlo simulation using this weight fraction range was used to assess risk. For
human cancer inhalation risk estimates were above 1 in 10,000 for both central tendency and high-end
exposures and dermal presented some risks as well. For environmental, this OES presents risk to the
aquatic environment at the high end but not at the central tendency by 1 day of exceedance. Although
risk does seem to decrease based on lower concentrations of TCEP being used in certain OESs, EPA has
not estimated a weight fraction value alone that would eliminate risk.
5.3.5.2 Consumer Risk Estimates
Lack of Weight Fraction Data
No safety data sheets (SDSs) were available for consumer products containing TCEP. Monitoring
literature and databases suggest that TCEP is used in consumer articles (e.g., fabric and textiles, home
furnishings, automobile foams, children's toys, and building materials such as insulation). Section
5.1.2.2 highlights the available information on the consumer COUs and relevant exposure scenarios.
EPA only had a few U.S. studies and databases (Castorina et al.. 2017; Fang et al.. 2013). including the
Ecology Washington Database (WSDE. 2023). which provides information on article weight fractions
for the consumer COUs. Where there were gaps, EPA utilized foreign data (Ionas et al.. 2014; Marklund
et al.. 2003; Ingerowski et al.. 2001) to help select values for product weight fraction data. EPA is
unclear on how relevant the foreign weight fraction data are for consumer articles used in the United
States. Moreover, one of these European studies (Ingerowski et al.. 2001) had a low-quality data
evaluation rating and was from the early 2000s. In addition, there are limitations in the data integrity in
the Washington State Database (WSDE. 2023). There is a possibility that a chemical could be a
contaminant rather than a component of the formulation of the consumer article. In addition, there are
some quality assurance and quality control issues with the database suggesting that it might be
unreliable.
Nevertheless, due to the paucity of information, EPA used low-quality information where higher quality
information was unavailable. In general, EPA has slight confidence in the building and construction
materials COUs (e.g., insulation and acoustic ceiling); slight-moderate confidence in the wood resin
products and foam seating and bedding products exposure scenarios; and moderate confidence in the
fabric and textile COUs (e.g., carpet back coating).
Page 370 of 638
-------
Complexity of Exposure Scenarios
The indoor air and indoor dust literature indicate that TCEP is present at higher values in indoor vs.
outdoor environments suggesting amplified exposures in the home. Uncertainties in the particle and gas
distribution (see Section 3.3.1.2.1) of TCEP builds further uncertainty on the reliability of direct
inhalation estimates vs. dust-mediated exposure via dermal absorption and oral ingestion.
SVOCs such as TCEP exhibit complex behaviors in the indoor environment. Shin et al. (2014) indicates
that TCEP has a relatively high emission rate compared to other semivolatile organic compounds. Shin
et al. (2014) observed that dust parameters such as removal rate from vacuuming, and dust loading onto
carpets and indoor furnishings are important variables that influence emission rates. CEM 3.2 does
incorporate defaults for cleaning frequency and cleaning efficiency from settled floor dust; however,
EPA was not able to obtain data on dust loading onto carpets when assessing the consumer COUs. The
uncertainties related to the behavior of TCEP in the indoor dust matrix further builds uncertainty into the
consumer risk estimates.
Model and Parameter Uncertainties
CEM 3.2 is a deterministic (rather than a population-based) model that provides point estimates of
TCEP exposure to population of interest. CEM is not equipped to model complex emission profiles or
activity patterns of residents other than those pre-populated within CEM. EPA used the CEM 3.2's
sensitivity mode to vary certain parameters to help understand which parameters influence the exposure
estimates. The initial concentration of SVOC in the article (a product of weight fraction and product
density) was the most important parameter for consumer modeling. Best judgments were used to
approximate product density of consumer articles where defaults were unavailable. The uncertainties in
the weight fraction and density information are reflected in EPA's overall confidence in consumer
modeling.
Dermal absorption parameter of fraction absorbed (Fabs) was estimated at 35.1 percent for all consumer
article scenarios from Abdallah et al. (2016). This value overrode the embedded CEM calculation for
dermal absorption. Estimates derived from the literature were of higher confidence then the CEM 3.2
calculated dermal absorption parameters. Nevertheless, there are uncertainties as to the applicability of
this one fraction absorbed value for all scenarios. Fraction absorbed can be a function of duration of
article or dust contact; however, because EPA was uncertain as to how often consumers, infants, and
children would wash their hands, EPA retained a conservative fraction absorbed value for the purposes
of consumer modeling.
Monitoring vs. Modeled Concentrations and Doses
The incongruence between modeled and measured concentrations and doses helps illustrate further
uncertainties in the consumer exposure assessment. Modeled indoor air concentrations for the
building/construction materials, insulation scenario (12.07 mg/m3) are six orders of magnitude higher
than the highest indoor air TCEP concentration observed in the United States (95th percentile of 35
ng/m3) (Dodson et al.. 2017). This discrepancy suggests major uncertainties in the insulation exposure
scenario.
The highest observed modeled dust intake in the reported modeled literature was 1.38 |ig/kg-day
reported for children at a kindergarten in Hong Kong (Deng et al.. 2018b). This value is within one to
two orders of magnitude of EPA's highest oral and dermal modeled intakes for children. EPA's highest
modeled oral intakes was 6.92><10~2 mg/kg-day (69.2 |ig/kg-day) for the foam toy block scenario. EPA's
highest observed dermal intakes via dermal absorption was 3,07/10 1 mg/kg-day (307 |ig/kg-day) for
Page 371 of 638
-------
the wood flooring scenario. These comparisons suggest that the oral and dermal intakes are more like
values reported in the literature than the modeled inhalation estimates.
Timeseries of Inhalation Exposure Estimates
CEM 3.0 estimates a chronic inhalation exposure by averaging the exposure over 365 days. Chronic
consumer inhalation exposures from TCEP containing articles are initially dominated by the gas phase
concentrations (due to off-gassing of TCEP). Figure 5-20 and Figure 5-21 display the time series air
concentrations for acoustic ceilings and wood flooring scenarios. After 4 weeks for the acoustic ceiling
scenario and 2 weeks for the wood flooring scenario, chronic consumer inhalation exposures are
dominated by the dust air concentrations. Chronic inhalation concentrations from insulation were
dominated by the gas phase concentrations; however, Figure 5-22 displays a precipitous drop in
concentration from the insulation article after the first few months.
Acoustic Ceiling Time Series Air Concentrations
100
o
U
u o.oooi
^ 0.00001
o
0.001
0.01
0.1
10
1
rHrHrtrHrHrHrHrHMMMMMMMmmnmmm
Mean Gas Phase Cone (mg/m3)
Dust Air Cone (mg/m3)
Days
TSP Air Cone (mg/m3)
— Abraded Particle Air Cone (mg/m3)
Figure 5-20. Consumer Modeling Time Series Results for Acoustic Ceilings
Page 372 of 638
-------
Wood Flooring Time Series Air Concentrations
10000
Days
at)
£
c
'o
!-h
c
(L)
O
c
o
u
00
o
—1
0.0001
0.00001
'-HfN^ri^i'Ot^ONO'-Hm^ri^t^OOONOCNm^r'Ot^OOO'-HCNmi^'O
¦Mean Gas Phase Cone (mg/m3) TSP Air Cone (mg/m3)
Dust Air Cone (mg/m3) ^—Abraded Particle Air Cone (mg/m3)
Figure 5-21. Consumer Modeling Time Series Results for Wood Flooring
Insulation Time Series Air Concentrations
1000
O0
c
o
!-h
c
(L)
O
c
o
u
00
o
—1
0.01
0.001
0.0001
0.00001
0.000001
(N^-i^it^OOON'-HCN^ri^'OOOON'-HCNmi^'OOOONOCNmi^'O
Days
Mean Gas Phase Cone (mg/m3)
TSP Air Cone (mg/m3)
Figure 5-22. Consumer Modeling Time Series Results for Insulation
Consumer articles containing TCEP are no longer manufactured in the United States. Consumers may
obtain new products containing TCEP only via import. Older articles in the home may have already
Page 373 of 638
-------
undergone off-gassing of TCEP; thus, there is uncertainty as to the relevance of continued inhalation
exposure from older consumer articles containing TCEP as much of the exposure may have already
occurred in the first few weeks.
Risk Estimates for Conservative Scenarios
EPA did not utilize a range of estimates to model a central tendency and high-end for consumer
exposures. Detection frequencies of TCEP were low for various consumer products in the Washington
State Database and accompanying monitoring data, and rather than utilize a central tendency (that
potentially was below realistic detection limits), EPA selected plausible worst-case values for weight
fractions. Due to this approach, EPA has more confidence in scenarios that did not exhibit risk than
scenarios that exhibited risk.
Sensitivity for Minimum Weight Fraction Values
Table 5-68 indicates that EPA has moderate confidence in the non-cancer MOEs for the Fabric and
textile products scenario - textile for children's outdoor play structures, Foam seating and bedding
products - foam auto, and building/construction materials - wood and engineered wood products -
wood resin composites - wooden TV stand scenarios. EPA used CEM 3.2's batch mode, to vary weight
fractions two to three orders of magnitude.
The wooden TV stand scenario had the highest weight fraction (3%) of the scenarios with moderate
confidence. When keeping all other parameters constant, varying the weight fraction to 1 percent no
longer resulted in dermal risk for infants and children, and no longer resulted in ingestion risk for
children. However, even after varying the weight fraction to 1 percent, 0.1 percent, and 0.01 percent,
EPA still saw inhalation and ingestion risk for Infants.
EPA also varied weight fractions for the textile for children's outdoor play structures scenario. This
scenario had an initial weight fraction of 1.3 percent and moderate confidence. When keeping all other
parameters constant, and varying the weight fraction to 1 percent, EPA still saw inhalation risk to
children and infants, and ingestion risk to infants. When varying the weight fraction to 0.1 percent, EPA
no longer saw inhalation risk for any lifestages, but still saw ingestion risk to infants. Even after varying
the weight fraction to 0.01 percent EPA still saw ingestion risk for Infants.
EPA varied weight factions for the automobile foam scenario (0.74 percent weight fraction and
moderate confidence). Even after varying the weight fraction to 0.1 percent and 0.01 percent, EPA still
saw ingestion risk for Infants.
These results indicate that TCEP weight fraction does not appear to be a good predictor of risk from
TCEP ingestion and inhalation exposure to Infants. More information on this sensitivity analysis can be
found in the (U.S. EPA 2024b)
5.3.5.3 General Population Risk Estimates
Location Information
Due to the lack of reasonably available site-specific information, the exposures assessment relied on
assumptions for location specific model inputs. This lack of data results in uncertainties surrounding
these location specific parameters (e.g., flow parameters and meteorological data). The AERMOD
Model included two meteorological conditions (Sioux Falls, South Dakota, for central tendency
meteorology and Lake Charles, Louisiana, for higher-end meteorology), in addition to different land
coverage scenarios (Suburban Forests and Oceans) to characterize potential amounts of annual TCEP
Page 374 of 638
-------
deposition to soil from air. It is unclear how relevant these meteorological conditions and land cover
scenarios are to TCEP facilities as there are no available site-specific information.
EPA modeled air concentrations and deposition fluxes at various distances from the hypothetical facility
releasing TCEP. The Agency selected various distances to calculate exposure doses and inhalation
concentrations for the general population (e.g., ambient air exposure to the general population, soil
dermal and oral intakes for children). In general, EPA has more confidence in risk estimates at further
distances from the hypothetical facility than risk estimates at closer distances. For example, EPA has
less confidence soil dermal exposure at 100 m of the facility than it does with soil dermal exposure at
1,000 m of the facility.
Due to the lack of reasonably available site-specific information for industrial and commercial releases
of TCEP, EPA could not estimate the proximity of general population residents to drinking water intake
locations. Drinking water estimates were calculated for non-diluted (i.e., drinking water intake locations
are at the site of the surface water release) conditions as a worst-case scenario. Drinking water estimates
were also calculate for diluted conditions by estimating the distance between intake location and the site
of release via drinking water intake information available for various SIC codes. EPA has more
confidence in these estimates as they represent a more plausible distance from which the general
population would receive their drinking water.
EPA Region 9 (2023) conducted a biological evaluation and habitat assessment for discharges to the
Pacific Ocean with the city of Los Angeles Hyperion Water Reclamation Plant and wastewater
reclamation systems. This study indicated average mass loadings of TCEP from 0.19 to 0.28 lb/day. In
addition, this study reported effluent concentrations of TCEP between 0.10 to 0.15 |ig/L after wet and
dry events in Fall of 2019. Furthermore, this study estimated annual average discharge rates of 230 to
236 MGD (millions of gallons per day) which equates to 871 to 893 MLD (millions of liters per day).
Due to the absence of information on specific sites and water bodies receiving TCEP water releases,
EPA utilized SIC codes to estimate flow rates. EPA used 50th percentile harmonic mean flow rates
ranging from 6.51 to 11.46 MLD. In the Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) -
Supplemental Information File: E-FAST Modeling Results (U.S. EPA. 2024g). EPA has included surface
water modeling estimates for the 90th percentile flow rates which range from 1,780 to 14,942 MLD).
The effluent flows presented in the analysis conducted by EPA Region 9 (2023) (871-893 MLD) are in
between the 50th percentile and 90th surface water flow rates utilized in EPA's surface water modeling.
Surface water flow rates are generally larger than wastewater flow rates. Although, EPA utilized the
more conservative 50th percentile harmonic mean flow (3.51-11.46) in its risk calculations, the flows
described in the EPA Region 9 (2023) assessment should be interpreted with caution as they are only
one data point compared to thousands of facilities that are sampled from the SIC codes. The approach
for calculating 50th percentile and 90th percentile flows are described in Appendix 1.2.2.
Results of the 90th percentile harmonic mean stream flows and central tendency estimates are presented
in Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File: E-FAST
Modeling Results (U.S. EPA. 2024g). These results indicate the critical role of receiving water flow as
an input in determining TCEP concentrations in surface water.
Monitoring vs. Modeled Concentrations and Doses
The incongruence between modeled and measured concentrations and doses helps illustrate further
uncertainties in the general population. WQP data on surface water TCEP concentrations is three to five
Page 375 of 638
-------
orders of magnitude lower than modeled surface water concentrations (see Sections 3.3.2.4 and 3.3.2.5).
TCEP fish tissue concentrations within the Great Lakes (Guo et al.. 2017b) are two to three orders of
magnitude lower than the TCEP tissue concentrations calculated using a whole organism BCF value
from another high-quality study (Arukwe et al.. 2018). Modeled soil concentrations were within one
order of magnitude of a single study from published literature (Mihailovic and Fries. 2012); however, it
is important to note that similarity with a single study is not enough to build confidence in the relevance
or accuracy of modeled results.
Complexity of Exposures Scenarios
The dermal absorption and ingestion from soil exposures scenarios require a complex understanding of
fate and transport of TCEP. Soil concentrations were calculated by modeling deposition fluxes of TCEP
at various distances from a hypothetical facility. Soil intakes were estimated for two exposures
scenarios—a child playing in mud and a child performing activities with soil. Parameters to calculate
these exposures, such as surface areas, absorption factors, and intake rates, were available in EPA's
Exposure Factors Handbook (U.S. EPA. 2017d); however, there is high uncertainty in the scenario due
to the multiple unknowns (e.g., hypothetical facility, hypothetical release estimate, unknown distance
between homes and facility).
Model and Parameter Uncertainties
An additional uncertainty for the general population and consumer assessment are model uncertainties.
VVWM-PSC allowed for the application of a standard, conservative, set of parameters and adjust for
physical-chemical properties of TCEP. For example, stream reach was set to represent a shallow
waterway with a width of 5 m and depth of 1 m. There are uncertainties on the applicability of this
shallow water body volume.
Ambient and drinking water estimates via VVWM-PSC and EFAST utilized a 0 percent drinking water
treatment removal efficiency (see Section F.2.5.3). Although TCEP has been shown to be recalcitrant to
removal treatment processes, EPA is uncertain whether advanced treatment methods can remove TCEP
from water.
For AERMOD, EPA specified deposition parameters for such as the fraction of gas vs. particle phase,
diffusivity in air, diffusivity in water, and the MMAD. Further sensitivity analysis can illustrate the
effects these parameters have on the deposition fluxes. Conflicting information in the peer-reviewed
literature creates uncertainties on the appropriate values of these parameters. Okeme (2018) has
described the complexities associated with the gas and particle partitioning of TCEP and has suggested
reported high concentrations of TCEP in particulates may be a result of sampling artifact (see Section
3.3.1.2.1).
A major uncertainty in fish ingestion exposure estimates was the selection of BAF values. Appendix
F.2.6 provides a review of BAFs found in the literature. The BAF of 2,198 for walleye (Sander vitreus)
from Guo et al. (2017a) was initially selected as a representative study of the U.S. population as it
sampled surface water and fish tissue concentrations in the Great Lakes. Walleye also represent a cool-
water top predator that serves as an important food fish. This species potentially preys on secondary and
tertiary consumers; however, it is uncertain what localized conditions affect BAF values within Guo et
al. (2017a). Furthermore, the surface water concentration and fish tissue concentrations were collected in
different years, thus it is difficult to hypothesize if TCEP surface water concentrations at the time of
sample collection influenced BAF values. A possible explanation for the resulting high oral risk
estimates could be an issue specific to BAFs for walleye (Sander vitreusj within the selected study Guo
et al. (2017a).
Page 376 of 638
-------
Uncertainties in Production Volume and Leachate Concentrations for Disposals Analysis
As demonstrated in Figure 1-3, the production volume of TCEP has decreased over the past decade.
Current production volume levels are difficult to predict. EPA conducted a bounding analysis to
estimate groundwater concentrations resulting from the disposal of TCEP to landfills utilizing the DRAS
software (see Section 3.3.3.8). The bounding analysis varied the production volume four orders of
magnitude to account for past disposal practices. In addition, EPA varied the leachate concentrations,
bounding the top of the range by TCEP's solubility (see Section 3.3.3.8).
Although efforts were made to estimate the potential migration of TCEP to groundwater from the
disposal of TCEP containing wastes, there are considerable uncertainties in this assessment approach.
Uncertainties in the loading rate, the leachate concentration and the absence of site-specific information
make it difficult to characterize this exposure scenario.
When estimating drinking water risk using the estimated groundwater concentrations from the DRAS
analysis, EPA only found lifetime cancer risk when using a leachate concentration of 1,000 mg/L and
production volume above 250,000 lb. No lifetime cancer risk was observed for a leachate concentration
of 100 mg/L and a production volume of 2,500,000 lb. TCEP's solubility is 7,820 mg/L (see Table 2-1).
EPA believes that 2,500 lb is the most suitable production volume for current uses of TCEP, and the
highest reported literature value of TCEP in leachate concentrations was 0.177 mg/L (Masoner et al..
2014a).
Therefore, this analysis suggests that even with the uncertainties of no site-specific information,
assuming groundwater concentrations in place of drinking water concentrations, and assuming the
general population living in proximity to poorly managed landfills, production volumes would have to
be three orders of magnitude higher than current levels, and leachate concentrations would have to be
four orders of magnitude higher than the maximum observed in the monitoring literature for TCEP to
display risk via drinking water ingestion.
Risk Estimates for Conservative Scenarios
To help characterize risk EPA uses a range of central tendency and high-end estimates, as well as
varying scenarios. EPA has more confidence in a risk estimate when risk is observed using conservative
assumptions. In addition, EPA has more confidence in risk estimates when risk is not observed using
fewer conservative assumptions. No risk observed with conservative parameters can build confidence
that the OES/COU is not a risk to consumers or the general population. For example, drinking water
risks were estimated for drinking water, diluted drinking water, incidental ingestion via swimming and
drinking water contamination from landfill leachate. None of these scenarios resulted in chronic oral
risk. Lifetime cancer risks were found for a few OESs (Incorporation into 1-part and 2-part reactive
paints and coatings, Commercial use of paints and coatings, and Processing of 2-part resin articles);
however, when adjusting for dilution to drinking water intake locations, these OESs no longer show
lifetime cancer risk.
Due to the uncertainties in the BAF for walleye, EPA considered BAF values from all reviewed studies
to capture a range conditions (see Section 2.12.2). Liu et al. (2019a) measured BAFs for multiple aquatic
species in China and reported the lowest value of 109 to 202 L/kg for mud carp (Cirrhimts molitorella).
Samples were collected from an e-waste polluted pond in South China. Risk estimates using this lowest
BAF value (109 L/kg) still resulted in risks for fish consumption (see Table 5-70). Lastly, EPA's
modeled surface water concentrations are generally several magnitudes higher than measured
concentrations, thus resultant fish tissue concentrations and doses are high regardless of BAF. However,
the Agency still relied on modeled data because of the paucity of measured data.
Page 377 of 638
-------
5.3.5.4 Hazard Values
EPA has moderate confidence in all hazard values used to modeled risks from TCEP. Although
additional toxicity data can always help to refine the risk evaluation, EPA believes the available human
studies and oral animal toxicity studies address the relevant endpoints for TCEP (e.g., neurotoxicity,
reproductive toxicity, developmental toxicity, other repeated-dose endpoints, carcinogenicity). There are
uncertainties that are common to all values. EPA identified several epidemiology studies that added to
the weight of scientific evidence, but the studies were not considered useful for dose-response analysis.
Therefore, the Agency used TCEP values from oral toxicity studies in animals, and these values required
extrapolation to inhalation and dermal hazard values. The impact of these uncertainties on the direction
of risk (under- or overprediction) is unknown. Additional uncertainties specific to individual hazard
values are described below, with details presented in Section 5.2.6.
Acute HED and HEC
Based on the weight of scientific evidence analysis of the reasonably available toxicity studies from
animals, the key acute exposure effect is neurotoxicity. EPA identified a POD from high-quality acute
animal toxicity study to calculate risks for acute exposure scenarios for TCEP. Tilson et al. (1990)
identified neurotoxicity in female rats, and EPA concluded that these types of effects are likely to be
caused by TCEP. EPA did not identify human data or other animal toxicity data using acute exposure
durations, and there is uncertainty because the POD does not account for all the effects associated with
acute exposure.
Intermediate/Chronic HED and HEC
EPA concluded that reproductive toxicity in humans is likely to be caused by TCEP and identified a
high-quality 35-day study in adolescent male mice that identified decreases in seminiferous tubule
numbers as the non-cancer POD for both intermediate and chronic exposure scenarios (Chen et al..
2015a). The observed effect is adverse and fertility due to male reproductive effects is known to be
sensitive in humans. Using Chen et al. (2015a) for the POD is expected to be protective of other hazards
(e.g., neurotoxicity) for these exposure durations. There is uncertainty about the precision of the doses
because Chen et al. (2015a) is a dietary study and the authors did not state the amount of food
consumed. Using a 35-day toxicity study for chronic exposure durations adds some uncertainty (e.g., the
POD for the same effect may be lower after chronic exposure) but based on the weight of scientific
evidence for other studies with male reproductive toxicity at higher doses and limited data from an
unobtainable inhalation study that identified effects related to male reproductive toxicity and fertility,
EPA believes the use of this study is relevant for the chronic duration.
Cancer CSF and IUR
Integrating evidence from humans, animals, and mechanistic studies resulted in a conclusion that TCEP
is likely to cause cancer in humans under relevant exposure circumstances. EPA used a sensitive
endpoint, kidney tumors in male rats, from a high-quality study (NTP. 1991b) to estimate cancer risks
from exposure to TCEP. The increased incidence of renal tubule adenomas and carcinomas is
considered adverse, relevant to humans, and representative of a continuum of benign to malignant
tumors. Increased incidence of tumors was identified in one epidemiological study that identified an
association between TCEP and thyroid tumors (Hoffman et al.. 2017). Because NTP (1991b) identified
primarily benign kidney tumors (adenomas), the incidence of malignant tumors is less certain. However,
humans may be more sensitive and develop malignancies sooner than rats. Use of linear low dose
extrapolation is also uncertain because direct mutagenicity is not likely to be the predominant MO A;
thus, risks may be overpredicted using linear low dose extrapolation. Use of only kidney tumors could
result in some underestimation of risk.
Page 378 of 638
-------
Table 5-82. Overall Confidence for Acute, Intermediate, and Chronic Human Health Non-cancer Risk Characterization for COUs
Resulting in Risks" h
cou
Route/Exposed
Group
Exposure
Confidence
Hazard
Confidence
Risk
Characterization
Confidence
Life Cycle
Stage
Category
Subcategory
Occupational
Manufacturing
Import
Import
Dermal/Worker
++
++
Moderate
Processing
Processing - Incorporation
into formulation, mixture, or
reaction product
Paint and coating
manufacturing
Dermal/Worker
++
++
Moderate
Processing - Incorporation
into formulation, mixture, or
reaction product
Polymers used in
aerospace equipment and
products
Dermal/Worker
++
++
Moderate
Processing - Incorporation
into article
Aerospace equipment and
products and automotive
articles and replacement
parts containing TCEP
Dermal/Worker
++
++
Moderate
Commercial
Use
Paints and coatings
Paints and coatings
Inhalation/W orker
++
++
Moderate
Inhalation/ONU
+
++
Slight
Dermal/Worker
++
++
Moderate
Laboratory chemicals
Laboratory chemical
Dermal/Worker
++
++
Moderate
Consumer
Consumer Use
Paints and coatings
Paints and coatings
N/A
N/A
++
N/A
Furnishing, cleaning,
treatment/care products
Fabric and textile products
Oral
++
++
Moderate
Inhalation
++
++
Moderate
Furnishing, cleaning,
treatment/care products
Foam seating and bedding
products
Oral
++
++
Moderate
Construction, paint,
electrical, and metal
products
Building/construction
materials
Inhalation
+
++
Slight
Oral
++
++
Moderate
Page 379 of 638
-------
cou
Route/Exposed
Group
Exposure
Confidence
Hazard
Confidence
Risk
Characterization
Confidence
Life Cycle
Stage
Category
Subcategory
Construction, paint,
electrical, and metal
products
Building/construction
materials - Wood and
engineered wood products -
Wood resin composites
Inhalation
++
++
Moderate
Dermal
++
++
Moderate
Disposal
Disposal
Disposal
N/A
N/A
++
N/A
General population exposures
Manufacturing
Import
Import
Oral
+
++
Slight
Processing
Processing - Incorporation
into formulation, mixture, or
reaction product
Paint and coating
manufacturing
Oral
++
++
Moderate
Processing - Incorporation
into formulation, mixture, or
reaction product
Polymers used in aerospace
equipment and products
Oral
++
++
Moderate
Processing - Incorporation
into article
Aerospace equipment and
products and automotive
articles and replacement
parts containing TCEP
Oral
+
++
Slight
Commercial
Use
Paints and coatings
Paints and coatings
Oral
++
++
Moderate
Dermal
++
++
Moderate
Laboratory chemicals
Laboratory chemical
Oral
+
++
Slight
11 This table identifies COUs that have any non-cancer risk (acute, intermediate, or chronic) and the route associated with the risk.
b Intermediate risks were evaluated for workers only, not consumers or the general population.
Page 380 of 638
-------
Table 5-83. TCEP Evidence Table Summarizing Overall Confidence for Lifetime Human Health Cancer Risk Characterization for
CPUs Resulting in Risks
COUs
Route/Exposed
Group
Exposure
Confidence
Hazard
Confidence
Risk
Characterization
Confidence
Life Cycle
Stage
Category
Subcategory
Occupational
Manufacturing
Import
Import
Dermal/Worker
++
++
Moderate
Processing
Processing - Incorporation
into formulation, mixture, or
reaction product
Paint and coating
manufacturing
Dermal/Worker
++
++
Moderate
Processing - Incorporation
into formulation, mixture, or
reaction product
Polymers used in
aerospace equipment and
products
Dermal/Worker
++
++
Moderate
Processing - Incorporation
into article
Aerospace equipment and
products and automotive
articles and replacement
parts containing TCEP
Dermal/Worker
++
++
Moderate
Commercial
Use
Paints and coatings
Paints and coatings
Inhalation/W orker
++
++
Moderate
Inhalation/ONU
+
++
Slight
Dermal/Worker
++
++
Moderate
Laboratory chemicals
Laboratory chemical
Dermal/Worker
++
++
Moderate
Consumer
Consumer Use
Paints and coatings
Paints and coatings
N/A
N/A
++
N/A
Furnishing, cleaning,
treatment/care products
Fabric and textile products
Oral
++
++
Moderate
Inhalation
++
++
Moderate
Furnishing, cleaning,
treatment/care products
Foam seating and bedding
products
Oral
++
++
Moderate
Construction, paint, electrical,
and metal products
Building/construction
materials
Inhalation
+
++
Slight
Oral
++
++
Moderate
Page 381 of 638
-------
COUs
Route/Exposed
Group
Exposure
Confidence
Hazard
Confidence
Risk
Characterization
Confidence
Life Cycle
Stage
Category
Subcategory
Construction, paint, electrical,
and metal products
Building/construction
materials - Wood and
engineered wood products -
wood resin composites
Inhalation
++
++
Moderate
Dermal
++
++
Moderate
Consumer Use
Paints and coatings
Paints and coatings
N/A
N/A
++
N/A
Furnishing, cleaning,
treatment/care products
Fabric and textile products
Oral
++
++
Moderate
Inhalation
++
++
Moderate
Dermal
++
++
Moderate
Furnishing, cleaning,
treatment/care products
Foam seating and bedding
products
Oral
++
++
Moderate
Inhalation
++
++
Moderate
Dermal
++
++
Moderate
Construction, paint, electrical,
and metal products
Building/construction
materials
Oral
+
++
Slight
Inhalation
+
++
Slight
Dermal
+
++
Slight
Construction, paint, electrical,
and metal products
Building/construction
materials - Wood and
engineered wood products -
Wood resin composites
Oral
++
++
Moderate
Dermal
++
++
Moderate
Disposal
Disposal
Disposal
N/A
N/A
++
N/A
General population exposures
Manufacturing
Import
Import
Oral
+
++
Slight
Processing
Processing - Incorporation
into formulation, mixture, or
reaction product
Paint and coating
manufacturing
Oral
++
++
Moderate
Processing - Incorporation
into formulation, mixture, or
reaction product
Polymers used in aerospace
equipment and products
Oral
++
++
Moderate
Page 382 of 638
-------
COUs
Route/Exposed
Group
Exposure
Confidence
Hazard
Confidence
Risk
Characterization
Confidence
Life Cycle
Stage
Category
Subcategory
Processing - Incorporation
into article
Aerospace equipment and
products and automotive
articles and replacement
parts containing TCEP
Oral
+
++
Slight
Commercial
Use
Paints and coatings
Paints and coatings
Oral
++
++
Moderate
Inhalation
++
++
Moderate
Laboratory chemicals
Laboratory chemical
Oral
+
++
Slight
Page 383 of 638
-------
6 UNREASONABLE RISK DETERMINATION
TSCA section 6(b)(4) requires EPA to conduct a risk evaluation to determine whether a chemical
substance presents an unreasonable risk of injury to health or the environment, without consideration of
costs or other nonrisk factors, including an unreasonable risk to a PESS identified by EPA as relevant to
this risk evaluation, under the COUs.
EPA has determined that TCEP presents an unreasonable risk of injury to health and the environment
under the COUs. This unreasonable risk determination is based on the information in previous sections
of this risk evaluation and the appendices and supporting documents in accordance with TSCA section
6(b). It is also based on TSCA's best available science (TSCA section 26(h)), weight of scientific
evidence standards (TSCA section 26(i)), and relevant implementing regulations in 40 CFR part 702,
including the amendments to the procedures for chemical risk evaluation under TSCA finalized in May
2024 (89 FR 37028; May 3. 2024).
EPA will initiate risk management for TCEP by applying one or more of the requirements under TSCA
section 6(a) to the extent necessary so that TCEP no longer presents an unreasonable risk. The risk
management requirements will likely focus on the COUs significantly contributing to the unreasonable
risk. However, under TSCA section 6(a), EPA is not limited to regulating the specific COUs found to
significantly contribute to unreasonable risk and may select from among a suite of risk management
options related to manufacture, processing, distribution in commerce, commercial use, and disposal to
address the unreasonable risk. For instance, EPA may regulate upstream COUs (e.g., processing,
distribution in commerce) to address downstream COUs that significantly contribute to unreasonable
risk (e.g., consumer use)—even if the upstream COUs are not significant contributors to the
unreasonable risk. The Agency would also consider whether such risk may be prevented or reduced to a
sufficient extent by action taken under another federal law, such that referral to another agency under
TSCA section 9(a) or use of another EPA-administered authority to protect against such risk pursuant to
TSCA section 9(b), as appropriate.
The COUs evaluated for TCEP are listed in Table 1-1. The following COUs significantly contribute to
the unreasonable risk:
• Manufacturing (Import);
• Processing - Incorporation into formulation, mixture, or reaction product - Paint and coating
manufacturing;
• Processing - Incorporation into formulation, mixture, or reaction product - Polymers used in
aerospace equipment and products;
• Processing - Incorporation into article - Aerospace equipment and products and automotive
articles and replacement parts containing TCEP;
• Industrial use - Paints and coatings;
• Commercial use - Paints and coatings;
• Commercial use - Laboratory chemicals;
• Consumer use - Furnishing, cleaning, treatment/care products - Fabric and textile products;
• Consumer use - Furnishing, cleaning, treatment/care products - Foam seating and bedding
products; and
• Consumer use - Construction, paint, electrical, and metal products - Building/construction
materials - Wood and engineered wood products - Wood resin composites.
The following COUs do not significantly contribute to the unreasonable risk:
Page 384 of 638
-------
• Processing - Recycling;
• Distribution in commerce;
• Industrial use - Other use - Aerospace equipment and products and automotive articles and
replacement parts containing TCEP;
• Commercial use - Other use - Aerospace equipment and products and automotive articles and
replacement parts containing TCEP;
• Commercial use - Furnishing, cleaning, treatment/care products - Fabric and textile products;
• Commercial use - Furnishing, cleaning, treatment/care products - Foam seating and bedding
products;
• Commercial use - Construction, paint, electrical, and metal products - Building/construction
materials - Insulation;
• Commercial use - Construction, paint, electrical, and metal products - Building/construction
materials - Wood and engineered wood products - Wood resin composites;
• Consumer use - Paints and coatings, including those found on automotive articles and
replacement parts;
• Consumer use - Construction, paint, electrical, and metal products - Building/construction
materials - Insulation; and
• Disposal.
Because TCEP production volumes and uses have declined, and no companies reported manufacture or
import of TCEP in the 2020 CDR, EPA had limited data available to evaluate certain COUs. In
determining whether COUs that the Agency had limited information significantly contributed to the
unreasonable risk of TCEP, EPA integrated reasonably available information in a qualitative risk
characterization using professional judgement of read-across evidence. The qualitative analyses are a
best estimate of what EPA expects given the weight of scientific evidence without overstating the
science. Environmental and human health risk characterizations for those COUs with limited data are in
Sections 4.3.6.2 and 5.3.2. of this risk evaluation. Additional explanation regarding the qualitative risk
characterizations and EPA's conclusion about whether the COU significantly contributes to
unreasonable risk are included in Sections 6.1.4, 6.1.5, and 6.1.6. The COUs that significantly contribute
to unreasonable risk from TCEP are based on risk estimates that assume a production volume of 2,500
lb, which EPA has estimated, based on the reasonably available data, is reflective of current domestic
TCEP use.
Whether EPA makes a determination of unreasonable risk for a particular chemical substance under
TSCA depends upon risk-related factors beyond exceedance of benchmarks, such as the endpoint under
consideration, the reversibility of effect, exposure-related considerations (e.g., duration, magnitude,
frequency of exposure, population exposed), and the confidence in the information used to inform the
hazard and exposure values. For COUs evaluated quantitatively, to determine if a COU contributed
significantly to unreasonable risk, EPA compared the risk estimates of the scenario used to evaluate the
COUs and considered whether the risk from the COU was best represented by the central tendency or
high-end risk estimates. Additionally, in the risk evaluation, the Agency describes the strength of the
scientific evidence supporting the human health and environmental assessments as robust, moderate, or
slight. Robust confidence suggests thorough understanding of the scientific evidence and uncertainties,
and the supporting weight of scientific evidence outweighs the uncertainties to the point where it is
unlikely that the uncertainties could have a significant effect on the exposure estimate. Moderate
confidence suggests some understanding of the scientific evidence and uncertainties, and the supporting
scientific evidence weighed against the uncertainties is reasonably adequate to characterize exposure
estimates. Slight confidence is assigned when the weight of scientific evidence may not be adequate to
characterize the scenario, and when the Agency is making the best scientific assessment possible in the
Page 385 of 638
-------
absence of complete information. This risk evaluation discusses important assumptions and key sources
of uncertainty in the risk characterization, and these are described in more detail in the respective weight
of scientific evidence conclusions sections for fate and transport, environmental release, environmental
exposures, environmental hazards, and human health hazards. It also includes overall confidence and
remaining uncertainties sections for human health and environmental risk characterizations.
In the TCEP unreasonable risk determination, EPA has considered risk estimates with an overall
confidence rating of slight, moderate, or robust, and the Agency considered COUs with indeterminate
exposures and COUs with limited reasonably available information. In general, EPA makes an
unreasonable risk determination based on risk estimates that have an overall confidence rating of
moderate or robust—because those confidence ratings indicate the scientific evidence is adequate to
characterize risk estimates despite uncertainties or is such that it is unlikely the uncertainties could have
a significant effect on the risk estimates (see Appendix F.2.3.1).
6.1 Unreasonable Risk to Human Health
Calculated risk estimates (MOEs or cancer risk estimates) can provide a risk profile of TCEP by
presenting a range of estimates for different health effects for different COUs. When characterizing the
risk to human health from occupational exposures during risk evaluation under TSCA, EPA conducts
baseline assessments of risk and makes its determination of unreasonable risk from a baseline scenario
that does not assume use of respiratory protection or other PPE.43 A calculated MOE that is less than the
benchmark MOE supports a determination of unreasonable risk of injury to health, based on non-cancer
effects. Similarly, a calculated cancer risk estimate that is greater than the cancer benchmark supports a
determination of unreasonable risk of injury to health from cancer. It is important to emphasize that
these calculated risk estimates alone are not bright-line indicators of unreasonable risk.
6.1.1 Populations and Exposures EPA Assessed to Determine Unreasonable Risk to
Human Health
EPA has evaluated risk to workers, including ONUs and male and female adolescents and adults (>16
years old); consumer users; the general population; and infants via human milk from exposed
individuals, using reasonably available monitoring and modeling data for inhalation, dermal, and
ingestion exposures, as applicable. EPA has evaluated risk from inhalation and dermal exposure of
TCEP to workers as well as inhalation exposures to ONUs. The Agency also has evaluated risk from
oral, dermal, and inhalation exposures to consumers. For the general population, EPA has evaluated risk
from (1) ingestion exposures via drinking water, incidental surface water ingestion, fish ingestion
(including subsistence fishers), and soil ingestion by children; (2) dermal exposures to swimmers and
children playing in the mud and other activities with soil; and (3) chronic inhalation exposure. For
infants consuming the human milk of exposed individuals, EPA has evaluated risk from milk ingestion
based on milk concentrations modeled for maternal exposures associated with occupational, consumer,
and general population COUs. Descriptions of the data used for human health exposure and human
health hazards are provided in Sections 5.1 and 5.2 of this risk evaluation. Uncertainties for overall
exposures and hazards are presented in Section 5.3.5 and are summarized in Table 5-82 and Table 5-83
and are considered in the unreasonable risk determination.
6.1.2 Summary of Human Health Effects
EPA has determined that the unreasonable risk presented by TCEP is due to
43 It should be noted that, in some cases, baseline conditions may reflect certain mitigation measures, such as engineering
controls, in instances where exposure estimates are based on monitoring data at facilities that have engineering controls in
place.
Page 386 of 638
-------
• Non-cancer effects and cancer in workers from dermal and inhalation exposures;
• Non-cancer effects and cancer in consumers from ingestion, dermal, and inhalation exposures;
and
• Non-cancer effects and cancer in the general population, including subsistence and Tribal fishers
from fish consumption (ingestion).
With respect to health endpoints upon which EPA has based this unreasonable risk determination, the
Agency has moderate overall confidence in the following PODs: (1) acute neurotoxicity, (2)
intermediate and chronic male reproductive effects, (3) acute and chronic kidney effects, and (4) kidney
cancer. Only likely evidence integration conclusions, or those PODs with moderate overall confidence,
were considered for dose-response. The confidence on the PODs is explained in Section 5.2.6 and
Appendix L. EPA's exposure and overall risk characterization confidence levels varied and are
summarized in Table 5-82 and Table 5-83.
The health risk estimates for workers, ONUs, consumers, the general population, and infants through the
milk pathway are presented in Section 5.3.2. For consumer and general population exposures, risk
estimates are provided in Section 5.3 of this risk evaluation only when MOEs were less than the
benchmark MOEs for non-cancer effects or when cancer risks exceeded benchmark risk levels of 1 in
1,000,000 (lxlO 6). A complete list of health risk estimates for consumers and the general population is
in the following supplemental files of the risk evaluation (see also Appendix C): Risk Evaluation for
Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File: Exposure E-FAST 2014
Surface Water Modeling Inputs, Flow Data, and General Population Exposure Estimates and Risk
Calculations (U.S. EPA. 2024g). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) -
Supplemental Information File: Exposure Air Concentration Risk Calculations (U.S. EPA. 2024i). and
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File:
Supplemental Information on Consumer Exposure Modeling Results for TCEP (U.S. EPA. 2024f).
6.1.3 Basis for EPA's Determination of Unreasonable Risk to Human Health
In developing the exposure and hazard assessments for TCEP, EPA has analyzed reasonably available
information to ascertain whether some human populations may have greater exposure and/or
susceptibility than the general population to the hazard posed by TCEP. For the TCEP risk evaluation,
EPA has accounted for the following PESS: infants exposed through human milk from exposed
individuals, children and male adolescents who use consumer articles that contain TCEP or are among
the exposed general population, subsistence fishers, Tribal populations, pregnant women, workers and
consumers who experience aggregate or sentinel exposures, people who live in fenceline communities
that are near facilities that emit TCEP, and firefighters (see Section 5.3.3, Table 5-79, and Appendix E).
EPA has slight confidence in the inputs to the calculation of the risk estimates for infants ingesting
human milk from exposed individuals and cannot determine that the human milk pathway significantly
contributes to the unreasonable risk of TCEP for any COU (see Section 5.1.3.4.7 and Appendix 1.5.5).
There are no COUs that showed higher risk estimates in the infants compared to the mothers; in fact,
some COUs resulted in maternal doses and risk estimates that are one to two orders of magnitudes
higher for the mothers than the infants. Therefore, EPA has moderate confidence that protecting the
mother will also protect the infant from exposure via human milk.
Risk estimates based on high-end exposure levels (e.g., 95th percentile) are generally intended to cover
individuals with sentinel exposure levels, whereas risk estimates at the central tendency exposure are
generally estimates of average or typical exposures. EPA has aggregated exposures across certain routes
for consumers and identified at least two COUs where aggregating exposures across routes resulted in
Page 387 of 638
-------
risk where there was not risk when considering a single route. EPA has not aggregated exposures across
consumer COUs, because each COU already presented chronic risk to consumers. Because risk to the
general population was driven by sentinel exposures via fish ingestion, EPA has not aggregated risk
across routes or exposure scenarios for this population. EPA has not characterized aggregate risk to
workers. There were no instances of aggregate lifetime risk for any COU where there was not already
risk from the COU from an individual route (see Section 5.3.4). The UF of 10 for human variability that
EPA has applied to MOEs accounts for increased susceptibility of populations, such as children and
elderly populations. EPA also generally relies on high-end exposure levels to make an unreasonable risk
determination to capture vulnerable populations that are expected to have higher exposures. The non-
cancer PODs are based on susceptible populations. The acute POD is based on effects observed during
pregnancy, and the intermediate and chronic PODs are based on reproductive effects observed in
adolescent males. For cancer, although there is likely to be variability in susceptibility across the human
population, EPA has not identified specific human groups that are expected to be more susceptible to
cancer following TCEP exposure. More information on how EPA characterized sentinel and aggregate
risks is provided in Section 5.3.4.
6.1.4 Workers
Based on the occupational risk estimates and related risk factors, EPA has determined that cancer and
non-cancer effects from worker dermal exposure to TCEP in occupational settings for all COUs with
quantified risk estimates and worker inhalation exposure to TCEP from commercial and industrial use of
paints and coatings significantly contribute to the unreasonable risk presented by TCEP. More
information on occupational risk estimates is in Section 5.3.2.1 of this risk evaluation. One COU,
Processing - Recycling, has a cancer risk estimate for the high end at the benchmark for workers due to
inhalation; however, the central tendency is below the benchmark. In general, TCEP manufacturing and
processing was phased out starting in the late 1980s or early 1990s in favor of other flame retardants or
flame-retardant formulations, and the overall TCEP releases from the Processing - Recycling COU are
expected to continue to be low]er over time. Therefore, EPA considered the central tendency as the best
way to represent the contribution to unreasonable risk from this COU and determined that Processing -
Recycling does not significantly contribute to the unreasonable risk of cancer to workers from inhalation
exposure to TCEP.
EPA has determined that firefighters may be at increased risk of TCEP exposures during structure fires;
therefore, exposures to firefighters contribute to the unreasonable risk presented by TCEP. However,
EPA was not able to identify which specific COUs or pathway of exposures lead to the elevated TCEP
exposures during firefighting activities.
EPA used a Fractional Absorption Model to estimate dermal exposure to TCEP in occupational settings.
The model assumes a single exposure event per day and does not address variability in exposure
duration and frequency. However, even with these uncertainties and limitations, EPA has considered the
weight of scientific evidence for dermal risk estimates generated by the model to be sufficient for
determining whether a COU significantly contributes to unreasonable risk. More information on EPA's
confidence in these risk estimates and the uncertainties associated with them can be found in Section
5.1.1.4 of this risk evaluation.
The following occupational COUs do not have quantitative risk estimates for workers. However, EPA
has qualitatively evaluated the COUs by integrating limited amounts of reasonably available information
using professional judgement of read-across evidence. The qualitative analyses are a best estimate of
what EPA expects given the weight of scientific evidence without overstating the science (see Section
5.3.2.1.2):
Page 388 of 638
-------
• Distribution in commerce: EPA expects TCEP to be transported in sealed containers from import
sites to downstream processing and use sites, or for final disposal of TCEP. EPA expects under
standard operating procedures that exposures and releases that could occur during distribution in
commerce would not lead to a significant contribution to the unreasonable risk presented by
TCEP.
• Commercial use - Furnishing, cleaning, treatment/care products - Fabric and textile products;
Commercial use - Furnishing, cleaning, treatment/care products - Foam seating and bedding
products; Commercial use - Construction, paint, electrical, and metal products -
Building/construction materials - Insulation; Commercial use - Construction, paint, electrical,
and metal products - Building/construction materials - Wood and engineered wood products -
Wood resin composites: TCEP was used for these purposes in the past, but these COUs were
phased out beginning in the late 1980s or early 1990s and replaced by other flame retardants or
flame-retardant formulations. EPA assumes that any of these products still used commercially
represent a fraction of the overall amount of TCEP previously used for these purposes. Although
there might be some workers that are exposed to those older products, the consumer assessment
for these articles resulted in no indication of consumer risk for inhalation, ingestion, or dermal
exposures for adults for these COUs. Therefore, EPA has moderate confidence that under similar
exposure durations and exposure frequencies, the Agency expects these articles to pose no
commercial risk for inhalation, ingestion, or dermal exposures to commercial workers who use
articles in a similar fashion to consumers for these COUs. Thus, these COUs do not significantly
contribute to the unreasonable risk presented by TCEP.
• Disposal: For the commercial uses that have been phased out, any currently used products that
contain TCEP are expected to be disposed in landfills but will represent just a fraction of
previous amounts from when TCEP was used more widely. And similarly, only a small portion
of e-waste is expected to contain TCEP. However, although some releases and exposures could
occur during the disposal of the wide variety of items into which TCEP has been incorporated,
these are expected to be minimal and dispersed, and not expected to significantly contribute to
the unreasonable risk presented by TCEP.
6.1.5 Consumers
Based on the consumer risk estimates and related risk factors, EPA has found that the Consumer use -
Furnishing, cleaning, treatment/care products - Fabric and textile products COU significantly
contributes to lifetime cancer risks from inhalation and oral exposures. In addition, the Consumer use -
Furnishing, cleaning, treatment/care products - Foam seating and bedding products COU significantly
contributes to lifetime cancer risk due to dermal and oral exposures. The Consumer use - Construction,
paint, electrical, and metal products - Building/construction materials - Wood and engineered wood
products - Wood resin composites COU significantly contributes to the lifetime cancer risk from
inhalation, dermal, and oral exposures. The cancer risk estimates represent exposures at younger life
stages that may significantly contribute to the lifetime cancer risk.
In addition, EPA has found that non-cancer effects to infants through age 2 from ingestion of dust and
mouthing of articles covered by the Consumer use - Furnishing, cleaning, treatment/care products -
Fabric and textile products and foam seating and bedding products COUs, as well as from ingestion of
dust contaminated with TCEP from other articles in the home covered by the Consumer use -
construction, paint, electrical, and metal products - Building/construction materials - Wood and
engineered wood products - Wood resin composites consumer COU, significantly contribute to the
unreasonable risk presented by TCEP.
Page 389 of 638
-------
Additionally, dermal contact with TCEP from the Consumer use - Construction, paint, electrical, and
metal products - Building/construction materials - Wood and engineered wood products - Wood resin
composites COU significantly contributes to chronic risk for infants and children.
For adults, risk estimates from acute inhalation exposures to TCEP are below the MOE for two COUs:
Consumer use - Furnishing, cleaning, treatment/care products - Fabric and textile products and
Consumer use - Construction, paint, electrical, and metal products - Building/construction materials -
Wood and engineered wood products - Wood resin composites. Inhalation risks from these COUs
primarily occur within the first few weeks after the article is produced or painted due to off-gassing of
TCEP. Because of this, EPA has not anticipated a significant contribution to the unreasonable risk via
inhalation of TCEP from TCEP-containing products that have already been in commerce longer than the
off-gassing period, but the Agency has considered acute inhalation exposures from newer or imported
products containing TCEP. In the case of the Consumer use - Furnishing, cleaning, treatment/care
products - Fabric and textile products COU, TCEP incorporation into those consumer articles was
mostly phased out beginning in the late 1980s or early 1990s and replaced by other flame retardants or
flame-retardant formulations. EPA has concluded that this COU does not significantly contribute to the
unreasonable risk of TCEP due to acute inhalation by adults from older articles in the home that may
have already undergone off-gassing of TCEP; thus, the Agency does not expect acute inhalation
exposure to adults from consumer fabric and textile articles containing TCEP unless they are newer
articles (i.e., imported into the United States for use by infants and young children. EPA has reasonably
available information that imported fabric and textile articles expected to be used by individuals in this
age group could contain TCEP (Section 5.1.1.2). EPA is concluding that the fabric and textile products
COU significantly contributes to the unreasonable risk of TCEP due to acute inhalation by adults who
are exposed to newer, imported articles containing TCEP. Additionally, in the case of the Consumer use
- Construction, paint, electrical, and metal products - Building/construction materials - Wood and
engineered wood products - Wood resin composites, consumers may obtain new wood resin composite
products containing TCEP by importing them, therefore, EPA has concluded that this COU significantly
contributes to the unreasonable risk of TCEP due to acute inhalation by adults.
Two consumer COUs, Consumer use - Construction, paint, electrical, and metal products -
Building/construction materials - Insulation and Consumer use - Paints and coatings, including those
found on automotive articles and replacement parts, were found to not significantly contribute to the
unreasonable risk of TCEP because of the Agency's slight confidence in their risk estimates. Further
discussion can be found below and in Sections 5.3.2.2.1 and 5.3.2.2.2.
EPA's overall confidence in the acute, intermediate, and chronic consumer inhalation, ingestion, and
dermal exposure risk estimates ranges from slight to moderate—although only those risk estimates with
overall confidence of moderate or robust were considered in the unreasonable risk determination. More
information on the consumer analysis can be found in Sections 3.2.2, 3.4, 5.1.2, and 5.3.2.2 of the risk
evaluation.
The following consumer COU and associated disposal do not have quantitative risk estimates. However,
EPA has qualitatively evaluated the COUs by integrating limited amounts of reasonably available
information using professional judgement of read-across evidence. The qualitative analyses are a best
estimate of what EPA expects given the weight of scientific evidence without overstating the science.
The qualitative analyses are a best estimate of what the Agency expects given the weight of scientific
evidence without overstating the science (see Section 5.3.2.2.2).
• Consumer use - Paints and coatings, including those found on automotive articles and
replacement parts: Consumers are unlikely to obtain TCEP containing paints and coatings,
including those found on automotive articles and replacement parts, because the domestic retail
Page 390 of 638
-------
production and manufacturing of TCEP containing paints and coatings has ceased and TCEP
containing paints and coatings represent a small fraction of the paints and coatings on the market.
For consumer application of paints and coatings, EPA has determined that it is not reasonably
foreseen for consumers to obtain TCEP containing paints and coatings products that are available
for commercial applications and therefore does not expect exposure to consumers from the
application of TCEP containing paints and coatings. For consumer use of articles containing
dried paints and coatings, including those found on automotive articles and replacement parts,
EPA expects the exposure scenario to mirror the other consumer use articles (e.g., wood resin
articles) scenarios assessed in Section 5.3.2.2.1. EPA's consumer analysis for articles containing
TCEP resulted in no chronic inhalation, ingestion, or dermal risk for adults for the COUs with
moderate confidence. However, the consumer analysis did reveal chronic dermal risk for wood
resin composites and ingestion risk for multiple articles for infants and children. Although the
analysis revealed dermal and ingestion risk to infants and children from the use of articles
containing dried paints and coatings due to its similarity to the other consumer article scenarios
(e.g., wood resin articles), EPA's confidence in the analysis of the consumer risk from articles
with TCEP containing paint is slight. Therefore, EPA has determined this COU does not
significantly contribute to the unreasonable risk to consumers presented by TCEP.
• Disposal : Consumers may be exposed to articles containing TCEP during disposal and the
handling of waste. The removal of articles in DIY scenarios may lead to direct contact with
articles and the dust generated from the articles. EPA believes that the monitoring data found for
the commercial COU of e-waste recycling would represent similar exposures that could occur
during the removal and/or disposal of other articles containing TCEP. Risk to workers was not
found during these activities and therefore it is not expected that risk would be found in a DIY
scenario involving the removal and/or disposal of TCEP containing articles. Therefore, this COU
does not significantly contribute to the unreasonable risk to consumers presented by TCEP.
6.1.6 General Population
EPA has identified the following exposure routes as significantly contributing to the unreasonable risk
of TCEP for the following sub-populations:
Fish Ingestion
Based on the risk estimates for the general population, including subsistence fishers44, Tribal fishers, and
other related risk factors, EPA has determined that fish ingestion by the general population, including
subsistence fishers and Tribal fishers, for all COUs evaluated quantitatively contribute to the
unreasonable risk due to cancer. One COU significantly contributes to the unreasonable risk due to
chronic non-cancer effects due to general population fish ingestion. Additionally, EPA has determined
that three COUs significantly contribute to the unreasonable risk due to acute non-cancer effects for
subsistence fishers in the general population, and four COUs significantly contribute to the unreasonable
risk due to chronic non-cancer effects for subsistence fishers in the general population.
To make a determination of unreasonable risk based on fish consumption, EPA used a BAF of 109 L/kg
and an ingestion rate of 22.2 g/day (142.4 g/day for subsistence fishers) for adults aged 16 to less than
70 years to calculate risk estimates (see Section 5.1.3.4.3 and Table Apx 1-2). EPA's confidence in the
risk estimates using the BAF of 109 L/kg is moderate. Acute and chronic non-cancer risk estimates to
the general population for oral fish ingestion are provided in Table 5-71 and Table 5-73 of this risk
evaluation. Cancer risk estimates for oral fish ingestion are in Table 5-75 of this risk evaluation.
44 Subsistence fishers represent a PESS group for TCEP due to their increased exposure via fish ingestion (142.4 g/day
compared to a high-end of 22.2 g/day for the general population).
Page 391 of 638
-------
EPA has not determined whether Tribal populations were exposed to fish containing TCEP; however,
the Agency estimated Tribal ingestion rates based on available information, resulting in all COUs
quantitatively evaluated significantly contributing to the unreasonable risk due to acute and chronic non-
cancer effects and cancer. EPA has also estimated risk based on fish consumption at heritage rates.
Heritage rates refer to those that existed prior to non-indigenous settlement on Tribal fishery resources,
as well as changes in culture and lifeways (see Section 5.1.3.4.4). Additionally, based on the risk
estimates for adults, EPA has estimated that TCEP presents unreasonable risk of acute and chronic non-
cancer effects and cancer for children aged 15 years or less who consume fish tissue contaminated with
TCEP due to their higher rate of ingestion per kg of body weight.
Additionally, EPA has evaluated the following sub-populations and routes of exposure but did not
identify any significant contribution to the unreasonable risk of TCEP from these routes:
Drinking Water and Incidental Surface Water Ingestion
EPA has estimated that ingestion of drinking water (diluted), drinking water from groundwater
contaminated with TCEP leaching from landfills, and incidental surface water ingestion during
swimming do not significantly contribute to the unreasonable risk of TCEP for any COU. Acute oral
non-cancer risk estimates for drinking water and drinking water (diluted) ingestion for any age group
{i.e., adults >21, youths 16-20, youths 11-15, children 6-10, and toddlers 1-5 years) are presented in
Table 5-70 of this risk evaluation. Chronic non-cancer risk estimates for drinking water and incidental
surface water ingestion are provided in Table 5-72; cancer risk estimates from drinking water are
presented in Table 5-74.
Soil Ingestion
EPA has estimated that chronic soil ingestion does not significantly contributes to the unreasonable risk
of TCEP for any COU. EPA's confidence in the risk estimates at 1,000 m is moderate. Chronic non-
cancer risk estimates for soil ingestion are presented in Table 5-72 of this risk evaluation.
Incidental Dermal from Swimming
EPA has estimated that incidental dermal exposure to an adult swimming does not significantly
contribute to the unreasonable risk of TCEP for any COU. Dermal acute and chronic non-cancer risk
estimates for swimming are provided in Table 5-76 of this risk evaluation. EPA's confidence in the risk
estimates is moderate.
Children '.s Dermal Exposure from Playing in Mud and Soil Activities
EPA has estimated that chronic dermal exposure to children 3 to 6 years old playing in mud and
conducting soil activities does not significantly contribute to the unreasonable risk of TCEP for any
COU. EPA's confidence in the risk estimates at 1,000 m is moderate. Dermal, chronic non-cancer risk
estimates for children playing in mud and soil activities are included in Table 5-76 of this risk
evaluation.
Inhalation
EPA has calculated risk estimates for one COU, Commercial use - Paints and coatings. Chronic
inhalation non-cancer risk estimates indicating no risk for even the very conservative distance of 10 m
are in Table 5-77. Cancer risk estimates do not indicate risk for the distance between 100 to 1,000 m.
Cancer inhalation risk estimates are presented in Table 5-78 of this risk evaluation.
The following COUs do not have quantitative risk estimates for the general population. However, EPA
has qualitatively evaluated the COU by integrating limited amounts of reasonably available information
Page 392 of 638
-------
using professional judgement of read-across evidence. The qualitative analyses are a best estimate of
what EPA expects given the weight of scientific evidence without overstating the science (see Section
5.3.2.3.2):
• Processing - Recycling: EPA has not found reasonably available data to quantify environmental
releases of TCEP from e-waste facilities. The total releases are expected to be low because TCEP
is not typically used in electronics. For commercial and consumer COUs evaluated qualitatively,
according to literature sources, TCEP was used for these commercial and consumer COUs in the
past, but manufacturing and processing was phased out starting in the late 1980s or early 1990s
in favor of other flame retardants or flame-retardant formulations. The Agency assumes that
commercial and consumer products with TCEP that are still in use, but are no longer
manufactured or processed, represents a fraction of the overall amount of TCEP previously used.
Therefore, TCEP releases from this COU (Processing - Recycling) is expected to be lower than
those associated with COUs already quantified in this risk evaluation, and this COU does not
significantly contribute to the unreasonable risk of TCEP.
• Distribution in commerce: TCEP production volumes have declined and recent reports (e.g., the
2020 CDR cycle) indicate that production volumes may be below reporting levels; therefore, the
precise volume is unknown. The general decline in production volume would logically lead to
decreased distribution into commerce. Therefore, exposure and risk would also likely have
declined with time. Exposure is possible from ongoing manufacturing, processing, industrial, and
commercial uses. Nevertheless, given that TCEP and/or TCEP containing products or articles are
expected to be transported in sealed containers or packages; EPA anticipates that exposure and
releases to the general population during distribution in commerce will be negligible and not lead
to a significant contribution to the unreasonable risk presented by TCEP.
• Commercial use - Furnishing, cleaning, treatment/care products - Fabric and textile products;
Commercial use - Furnishing, cleaning, treatment/care products - Foam seating and bedding
products; Commercial use - Construction, paint, electrical, and metal products -
building/construction materials - Insulation; and Commercial use - Construction, paint,
electrical, and metal products - Building/construction materials - Wood and engineered wood
products - Wood resin composites: These COUs were being phased out beginning in the late
1980s or early 1990s and replaced by other flame retardants or flame-retardant formulations.
EPA did not locate data to estimate (1) the amount of TCEP that was historically used in these
products, (2) the amounts of these products that have already reached the end of their service life,
or (3) the amounts of these products that have already been disposed of. Based on the years that
the phase-out occurred, many of these products are not likely to be in use because the end of
their service life was already reached (e.g., commercial roofing has an estimated lifespan of 17 to
20 years). EPA assumes that any of these products still used commercially represent a fraction of
the overall amount of TCEP previously used for these purposes. Therefore, releases to the
environment from these commercial uses would also represent only a fraction of previous release
amounts and these COUs do not significantly contribute to the unreasonable risk of TCEP.
• Industrial use - Other use - Aerospace equipment and products and automotive articles and
replacement parts containing TCEP; and Commercial use - Other use - Aerospace equipment
and products and automotive articles and replacement parts containing TCEP: After TCEP-
containing resins have cured within products that are installed, EPA expects TCEP releases and
dermal exposures will be limited by TCEP being entrained into the hardened polymer matrix.
During installation it is possible that very small levels of dust could be generated, these were
quantified in Table 5-67. and do not indicate risk to workers from inhalation nor do they indicate
the generation of significant dust releases occurring. Releases may occur via the mechanism of
Page 393 of 638
-------
blooming (volatilization from the cured resin surface) during the service life of the article, but
EPA expects that such releases during installation will be negligible (OECD. 2009; NICNAS.
2001). Installation of aerospace equipment and products would be installed without any type of
further processing of the article that would lead to potential releases (sanding, drilling, etc.).
Therefore, the potential risk to the general population from releases during installation of TCEP -
containing articles is low. EPA has concluded these COUs do not significantly contribute to the
unreasonable risk presented by TCEP.
• Disposal: Disposal is possible throughout the life cycle of TCEP and TCEP-containing products,
including waste treatment and disposal resulting from manufacturing, processing, commercial
and consumer uses. A robust discussion of the qualitative analyses done to estimate risk to the
general population from the disposal of TCEP and TCEP-containing products can be found in
Section 5.3.2.3.2. EPA qualitatively discussed releases to landfills or incinerators, surface water,
groundwater, and considered releases due to e-waste recycling, end-of-life disposal, demolition,
and down-the-drain releases. The Agency acknowledges that although some releases and
exposures could occur during the disposal of the wide variety of items that TCEP has been
incorporated into, these exposures are expected to be negligible, and not expected to significantly
contribute to the unreasonable risk presented by TCEP.
6.2 Unreasonable Risk to the Environment
Calculated RQs can provide a risk profile by presenting a range of estimates for different environmental
hazard effects for different COUs. An RQ equal to 1 indicates that the exposures are the same as the
concentration that causes effects. An RQ less than 1, when the exposure is less than the effect
concentration, generally indicates that there is not risk of injury to the environment that would support a
determination of unreasonable risk for the chemical substance. An RQ greater than 1, when the exposure
is greater than the effect concentration, generally indicates that there is risk of injury to the environment
that would support a determination of unreasonable risk for the chemical substance. Additionally, if a
chronic RQ is 1 or greater, the Agency evaluates whether the chronic RQ is 1 or greater for 30 days or
more based on the exposure period of the hazard toxicity tests before making a determination of
unreasonable risk.
6.2.1 Populations and Exposures EPA Assessed to Determine Unreasonable Risk to the
Environment
For aquatic organisms, EPA has evaluated exposures via surface water and sediment (including pore
water). For terrestrial organisms, the Agency has evaluated exposures via soil, air, and surface water.
EPA has assessed terrestrial organism exposures from air deposition of TCEP to soil. Additionally, the
Agency has estimated terrestrial organism exposures from trophic transfer of TCEP from soil and
surface water.
6.2.2 Summary of Environmental Effects
EPA has determined that all five COUs assessed quantitatively, and one COU assessed qualitatively,
significantly contribute to the unreasonable risk presented by TCEP due to chronic effects (mortality of
yellow catfish) using empirical fish data.
Risks to terrestrial organisms from air deposition to soil and from trophic transfer from the five COUs
quantitatively assessed do not significantly contribute to the unreasonable risk to the environment
presented by TCEP.
Page 394 of 638
-------
6.2.3 Basis for EPA's Determination of Unreasonable Risk of Injury to the Environment
Consistent with EPA's determination of unreasonable risk to human health, the RQ is not treated as a
bright-line and other risk-based factors may be considered (e.g., confidence in the hazard and exposure
characterization, duration, magnitude, uncertainty) for purposes of making an unreasonable risk
determination. TCEP is described as a "ubiquitous" contaminant because it is commonly found in
various environmental compartments such as outdoor air, surface water, drinking water, groundwater,
soil, sediment, biota, and precipitation all over the world (see Section 3). Additionally, TCEP is
persistent in water, soil, and sediment, and EPA has robust confidence that TCEP can undergo long-
range transport.
EPA has moderate confidence in the acute and chronic aquatic hazards and aquatic exposures
significantly contributing to unreasonable risk. Additionally, the Agency has robust to slight confidence
in the terrestrial exposures and hazards, which do not significantly contribute to unreasonable risk. EPA
has determined the terrestrial food web to be the driver of exposure and does not expect exposure to
TCEP via air or surface water to significantly contribute to unreasonable risk to terrestrial organisms.
Similarly, EPA does not expect exposure to TCEP via biosolids to significantly contribute to
unreasonable risk to the environment. The Agency's overall environmental risk characterization
confidence levels were varied and are summarized in Table 4-23 of this risk evaluation.
In making a determination of unreasonable risk, EPA has considered aggregating environmental
exposures for aquatic and terrestrial organisms but did not because the surface water and sediment
pathways for aquatic organisms and the soil pathway for terrestrial organisms were such significant
contributors to unreasonable risk (see Section 4.3.6.1).
The following COUs do not have quantitative risk estimates for the environment. However, EPA has
qualitatively evaluated the COUs by integrating limited amounts of reasonably available information
using professional judgement of read-across evidence. The qualitative analyses are a best estimate of
what EPA expects given the weight of scientific evidence without overstating the science (see Section
4.3.6.2):
• Processing - Recycling: As noted before, the total releases are expected to be low because TCEP
is not typically used in electronics. Therefore, TCEP releases for this COU are expected to be
lower than those associated with COUs with quantified environmental risk estimates for
terrestrial receptors, with the Recycling COU expected to have lower risk than the quantified
COUs. Thus, EPA does not expect this COU to significantly contribute to the unreasonable risk
presented by TCEP to the environment.
• Distribution in commerce: EPA expects TCEP to be transported in sealed containers from import
sites to downstream processing and use sites. Under standard operating procedures, the Agency
expects that environmental releases from sealed containers are not expected to occur.
Transportation of TCEP and TCEP containing articles for final disposal would be less than the
releases expected from disposal, which are expected to be minimal and dispersed; therefore,
distribution in commerce is not expected to significantly contribute to the unreasonable risk
presented by TCEP to the environment.
• Industrial use - Paints and coatings: Similar to the occupational exposures, EPA expects
modeled environmental releases to be similar to the OESs previously assessed for other industry
sectors. This means that that the risk estimates for this COU would be similar to the commercial
use of paints and coatings; therefore, this COU also significantly contributes to the unreasonable
risk presented by TCEP to the environment.
Page 395 of 638
-------
• Industrial use - Other use - Aerospace equipment and products and automotive articles and
replacement parts containing TCEP and Commercial use - Other use - Aerospace equipment and
products and automotive articles and replacement parts containing TCEP: EPA does not expect
significant releases to the environment to occur during the installation of TCEP-containing
aircraft, aerospace, or automotive articles into or onto the relevant transportation equipment.
After TCEP-containing resins have cured, EPA expects TCEP release will be limited by the
hardened polymer matrix. Therefore, the Agency does not expect these COUs to significantly
contribute to the unreasonable risk presented by TCEP to the environment.
• Commercial use - Furnishing, cleaning, treatment/care products - Fabric and textile products;
Commercial use - Furnishing, cleaning, treatment/care products - Foam seating and bedding
products; Commercial use - Construction, paint, electrical, and metal products -
Building/construction materials - Insulation; and Commercial use - Construction, paint,
electrical, and metal products - Building/construction materials - Wood and engineered wood
products - Wood resin composites: EPA has confirmed from literature sources that TCEP was
used for these purposes in past decades. However, these commercial uses began phasing out
beginning in the late 1980s or early 1990s and were replaced by other flame retardants or flame-
retardant formulations. However, because TCEP releases are expected to be lower relative to
other quantified scenarios, these commercial COUs would be expected to have lower risk than
the quantified COUs. Therefore, EPA does not expect these COUs to significantly contribute to
the unreasonable risk presented by TCEP to the environment.
• Consumer use - Furnishing, cleaning, treatment/care products - Fabric and textile products;
Consumer use - Furnishing, cleaning, treatment/care products - Foam seating and bedding
products; Consumer use - Construction, paint, electrical, and metal products -
Building/construction materials - Insulation; Consumer use - Construction, paint, electrical, and
metal products - Building/construction materials wood and engineered wood products - Wood
resin composites; and Consumer use - Paints and coatings, including those found on automotive
articles and replacement parts: Consumer releases to the environment are expected to be less than
occupational releases; wastewater concentrations from manufacturing, commercial and
processing COUs were shown to be significantly lower than the acute and chronic COCs
identified in Section 4.2. Therefore, EPA does not expect these COUs to significantly contribute
to the unreasonable risk presented by TCEP to the environment.
• Disposal: As noted before, although some environmental releases could occur during the
disposal of the wide variety of items that TCEP has been incorporated in to, these releases are
expected to be minimal and dispersed and are not expected to significantly contribute to the
unreasonable risk presented by TCEP to the environment.
6.3 Additional Information Regarding the Basis for the Unreasonable Risk
Determination
Table 6-1, Table 6-2, and Table 6-3 summarize the basis for this unreasonable risk determination of
injury to human health and the environment presented in this TCEP risk evaluation for those COUs with
a qualitative evaluation. In these tables, a checkmark (•/) indicates how the COU significantly
contributes to the unreasonable risk by identifying the type of effect (e.g., non-cancer and cancer for
human health; acute or chronic environmental effects) and the exposure route to the population or
receptor that results in such significant contribution. Not all COUs, exposure routes, or populations or
receptors evaluated are included in the tables. The tables only includes the relevant exposure route, or
the population or receptor that supports the conclusion that the COU significantly contributes to the
TCEP unreasonable risk determination. As explained in Section 6.2, for this unreasonable risk
Page 396 of 638
-------
determination, EPA has considered the effects of TCEP to human health at the central tendency and
high-end, as well as effects of TCEP to human health and the environment from the exposures
associated from the condition of use, risk estimates, and uncertainties in the analysis. See Section 5.3.2
for a summary of risk estimates.
Page 397 of 638
-------
Table 6-1. Supporting Basis for the Unreasonable Risk Determination for Human
health (Occupational CPUs)
cou
Population
Exposure
Route
Human Health Effects
Life Cycle
Stage
Category
Subcategory
Acute
Non-cancer
Intermediate
Non-cancer
Chronic
Non-cancer
Lifetime
Cancer
Manufacturing
Import
Import
Worker
Dermal
/
~
~
~
General Population
Fish Ingestion
N/A
~
General Population - Subsistence Fishers
Fish Ingestion
N/A
~
~
General Population - Tribal Fishers - Current IR
Fish Ingestion
N/A
~
~
General Population - Tribal Fishers - Heritage IR
Fish Ingestion
/
N/A
~
~
Processing
Incorporation
into
formulation,
mixture, or
reaction
product
Paint and coating
manufacturing
Worker
Dermal"
~
~
~
~
General Population
Fish Ingestion
N/A
~
General Population - Subsistence Fishers
Fish Ingestion
~
N/A
•/
~
General Population - Tribal Fishers - Current IR
Fish Ingestion
~
N/A
~
General Population - Tribal Fishers - Heritage IR
Fish Ingestion
~
N/A
~
Incorporation
into
formulation,
mixture, or
reaction
product
Polymers used in
aerospace equipment
and products
Worker
Dermal
~
~
•/
~
General Population
Fish Ingestion
N/A
~
General Population - Subsistence Fishers
Fish Ingestion
~
N/A
~
General Population - Tribal Fishers - Current IR
Fish Ingestion
~
N/A
•/
~
General Population - Tribal Fishers - Heritage IR
Fish Ingestion
~
N/A
~
Incorporation
into article
Aerospace equipment
products and
automotive articles and
replacement parts
containing TCEP
Worker
Dermal 4
~
~
¦/
~
Commercial
and Industrial
Use
Paints and
coatings
Paints and coatings
Worker
Inhalation 4
~
~
¦/
~
Dermal 4
~
~
~
General Population
Fish Ingestion
N/A
~
General Population - Subsistence Fishers
Fish Ingestion
~
N/A
¦/
~
General Population - Tribal Fishers - Current IR
Fish Ingestion
~
N/A
¦/
~
General Population - Tribal Fishers - Heritage IR
Fish Ingestion
~
N/A
~
~
Page 398 of 638
-------
cou
Population
Exposure
Route
Human Health Effects
Life Cycle
Stage
Category
Subcategory
Acute
Non-cancer
Intermediate
Non-cancer
Chronic
Non-cancer
Lifetime
Cancer
Commercial
Use
Laboratory
chemicals
Laboratory chemical
Worker
Dermal
y
y
V
y
General Population
Fish Ingestion
N/A
y
General Population - Subsistence Fishers
Fish Ingestion
N/A
y
General Population - Tribal Fishers - Current IR
Fish Ingestion
N/A
y
General Population - Tribal Fishers - Fleritage IR
Fish Ingestion
N/A
y
" The risk estimate exceeded is based on the most conservative OES (1 -part coatings).
4 The risk estimate exceeded is based on the most conservative OES (2-part coatings, 250-day).
Page 399 of 638
-------
Table 6-2. Supporting Basis for the Unreasonable Risk Determination for Human Health
Consumer CPUs)
cou
Population"
Exposure Route
for Non-cancer
Acute
Non-
cancer
Chronic
Non-
cancer
Lifetime
Cancer6
Life Cycle
Stage
Category
Subcategory
Consumer
Use
Furnishing,
cleaning,
treatment/care
products
Fabric and
textile
products
Adult
Inhalation
V
¦S (inhalation
and oral)
Infant
Ingestion - Dust
and Mouthing
•/
•/
Furnishing,
cleaning,
treatment/care
products
Foam seating
and bedding
products
Adult
No Non-cancer
Risk
¦S (dermal and
oral)
Infant
Ingestion - Dust
and Mouthing
V
V
Construction,
paint,
electrical, and
metal products
Building/
construction
materials -
Wood and
engineered
wood
products -
Wood resin
composites
Adult
Inhalation
Y
S (dermal,
inhalation,
and oral)
Child
Dermal
•/
Infant
Ingestion - Dust
V
V
Dermal
V
" "Child" represents ages 3-10 years, and "Infant" represents ages 0-2 years.
h Risk estimates considered represent lifetime cancer risk and includes exposures at younger life stages that may
significantly contribute to the overall cancer risk.
c Consumer use of paints and coatings does not have quantitative risk estimates; EPA has conducted a qualitative
assessment.
d Inhalation of vapors generated from application of paints and coatings containing TCEP.
e Ingestion of dust generated by dried paints and coatings containing TCEP.
Page 400 of 638
-------
Table 6-3. Supporting Basis for the Unreasonable Risk Determination for the Environment
cou
Population/
Receptor
Compartment
Environmental
Effects
Life Cycle Stage
Category
Subcategory
Acute
Chronic
Manufacturing
Import
Import
Aquatic
Surface water
•/
Sediment
V
Processing
Incorporation into
formulation, mixture, or
reaction product
Paint and coating
manufacturing
Aquatic
Surface water
V
Sediment
V
Incorporation into
formulation, mixture, or
reaction product
Polymers used in
aerospace equipment
and products
Aquatic
Surface water
•/
Sediment
V
Commercial and
Industrial Use
Paints and coatings
Paints and coatings
Aquatic
Surface water
•/
Sediment
•/
Commercial Use
Laboratory chemical
Laboratory chemicals
Aquatic
Surface water
V
Sediment
V
Page 401 of 638
-------
REFERENCES
Abafe. OA; Martincigh. BS. (2019). Concentrations, sources and human exposure implications of
organophosphate esters in indoor dust from South Africa. Chemosphere 230: 239-247.
http: //dx. doi. or g/10.1016/i. chemosphere .2019.04.175
Abdallah. MA-E; Pa war. G; Harrad. S. (2016). Human dermal absorption of chlorinated
organophosphate flame retardants; implications for human exposure. Toxicol Appl Pharmacol
291: 28-37. http://dx.doi.Org/10.1016/i.taap.2015.12.004
Abdollahi. A: Eng. A: Jantunen. LM; Ahrens. L; Shoeib. M: Parnis. JM: Harner. T. (2017).
Characterization of polyurethane foam (PUF) and sorbent impregnated PUF (SIP) disk passive
air samplers for measuring organophosphate flame retardants. Chemosphere 167: 212-219.
http ://dx. doi. org/10.1016/i. chemosphere.2016.09.111
Aceto US LLC. (2024). Comment submitted by Aceto US, LLC regarding the TCEP Risk Evaluation.
(EPA-HQ-OPPT-2018-0476-0056). Port Washington, NY.
https://www.regulations.gov/document/EPA-HQ-OPPT-2023-0265-0Q37
AIA. (2019). Comment submitted by David Hyde, Director, Environmental Policy, Aerospace Industries
Association (AIA) regarding TCEP. (EPA-HQ-OPPT-2018-0476-0006). Arlington, VA.
https://www.regulations.gov/comment/EPA-HO-OPPT-2018-0476-00Q6
AIA. (2020). EPA and Aerospace Industries Association (AIA) Meeting on TCEP. (EPA-HQ-OPPT-
2018-0476-0034). Washington, DC: U.S. Environmental Protection Agency.
https://www.regulations.gov/comment/EPA-HQ-OPPT-2018-0476-0034
AIA. (2021a). EPA and AIA Meeting on TCEP, TPP, and TBBPA. (EPA-HQ-OPPT-2018-0476-0039).
Washington, DC: U.S. Environmental Protection Agency.
https://www.regulations.gov/comment/EPA-HQ-OPPT-2018-0476-0039
AIA. (2021b). Meeting with Aerospace Industries Association (AIA) and EPA to discuss comments
submitted for TBBPA, TPP, and TCEP. (EPA-HQ-OPPT-2018-0476-0040). Washington, DC:
U.S. Environmental Protection Agency. https://www.regulations.gov/comment/EPA-HQ-OPPT-
2018-0476-0040
Alexander. GR. (1977). Food of vertebrate predators on trout waters in north central lower Michigan.
Mich Acad 10: 181-195.
Ali. N: Dirtu. AC: van Den Eede. N: Goosey. E; Harrad. S: Neels. H; T Mannetie. A: Coaklev. J:
Douwes. J: Covaci. A. (2012). Occurrence of alternative flame retardants in indoor dust from
New Zealand: Indoor sources and human exposure assessment. Chemosphere 88: 1276-1282.
http ://dx. doi. org/10.1016/i. chemosphere.2012.03.100
Alliance for Automotive Innovation. (2023a). Comment submitted by Alliance for Automotive
Innovation regarding flame retardants and Significant New Uses Rules for certain non-ongoing
uses. (EPA-HQ-OPPT-2023-0012-0010). Washington, DC.
https://www.regulations.gov/comment/EPA-HQ-OPPT-2023-0012-001Q
Alliance for Automotive Innovation. (2023b). Letter from the Alliance for Automotive Innovation to
Discuss Use of TCEP in Automobile Articles. (EPA-HQ-OPPT-2018-0476-0059). Washington,
DC. https://www.regulations.gov/comment/EPA-HO-OPPT-2018-0476-0Q59
Alonso-Amelot. ME. (2018). Milk as a natural product: foreign natural and anthropogenic organic
compounds in it. In R Atta ur (Ed.), Studies in Natural Products Chemistry, vol 56 (pp. 335-435).
Amsterdam, Netherlands: Elsevier. http://dx.doi.org/10.1016/B978-0-444-64058-l.0001Q-8
Alzualde. A: Behl. M; Sipes. NS: Hsieh. JH; Aldav. A: Tice. RR; Paules. RS: Muriana. A: Quevedo. C.
(2018). Toxicity profiling of flame retardants in zebrafish embryos using a battery of assays for
developmental toxicity, neurotoxicity, cardiotoxicity and hepatotoxicity toward human
relevance. Neurotoxicol Teratol 70: 40-50. http://dx.doi.Org/10.1016/i.ntt.2018.10.002
Page 402 of 638
-------
Andresen. JA; Muir. D; Ueno. D; Darling. C; Theobald. N; Bester. K. (2007). Emerging pollutants in the
North Sea in comparison to Lake Ontario, Canada, data. Environ Toxicol Chem 26: 1081-1089.
http://dx.doi.Org/10.1897/06-416R.l
Araki. A: Bamai. YA; Bastiaensen. M; Van den Eede. N: Kawai. T; Tsuboi. T; Miyashita. C: Itoh. S:
Goudarzi. H; Konno. S: Covaci. A: Kishi. R. (2020). Combined exposure to phthalate esters and
phosphate flame retardants and plasticizers and their associations with wheeze and allergy
symptoms among school children. Environ Res 183: 109212.
http://dx.doi. org/10.1016/i. envres .2020.109212
Arnot. J: Gobas. F. (2006). A review of bioconcentration factor (BCF) and bioaccumulation factor
(BAF) assessments for organic chemicals in aquatic organisms. Environ Rev 14: 257-297.
http://dx.doi.Org/10.l 139/A06-005
Arukwe. A: Carteny. CC: Eggen. T; Moder. M. (2018). Novel aspects of uptake patterns, metabolite
formation and toxicological responses in salmon exposed to the organophosphate esters-tris(2-
butoxyethyl)- and tris(2-chloroethyl) phosphate. Aquat Toxicol 196: 146-153.
http://dx.doi.Org/10.1016/i.aquatox.2018.01.014
ASTM. (2014). Standard guide for conducting acute toxicity tests on test materials with fishes,
macroinvertebrates, and amphibians. (E 29-96(2014)). West Conshohocken, PA.
https://compass.astm.org/document/?contentCode=ASTM%7CE0729-96R14%7Cen-US
Aston. LS: Noda. J: Seiber. JN; Reece. CA. (1996). Organophosphate flame retardants in needles of
Pinus ponderosa in the Sierra Nevada foothills. Bull Environ Contam Toxicol 57: 859-866.
http://dx.doi.org/10.1007/sQ01289900269
ATSDR. (2012). Toxicological profile for phosphate ester flame retardants. (PB2013-105796). Atlanta,
GA: U.S. Department of Health and Human Services, Public Health Service.
http://www.atsdr.cdc.gov/toxprofiles/tp202.pdf
Awkerman. JA: Raimondo. S: Jackson. CR; Barron. MG. (2014). Augmenting aquatic species
sensitivity distributions with interspecies toxicity estimation models. Environ Toxicol Chem 33:
688-695. http://dx.doi.org/10.1002/etc.2456
Awonaike. B; Lei. YD: Wania. F. (2021). Precipitation-induced transport and phase partitioning of
organophosphate esters (OPEs) in urban and rural watersheds. Environ Sci (Camb) 7: 2106-
2115. http://dx.doi.org/10.1039/DlEW00329A
Bacaloni. A: Cucci. F; Guarino. C: Nazzari. M; Samperi. R; Lagana. A. (2008). Occurrence of
organophosphorus flame retardant and plasticizers in three volcanic lakes of central Italy.
Environ Sci Technol 42: 1898-1903.
Barnes. KK; Christenson. SC: Kolpin. DW: Focazio. M; Furlong. ET; Zaugg. SD; Meyer. MT; Barber.
LB. (2004). Pharmaceuticals and other organic waste water contaminants within a leachate
plume downgradient of a municipal landfill. Ground Water Monit Remediat 24: 119-126.
http://dx.doi.org/10.1111/i. 1745-6592.2004.tb00720.x
Barnthouse. LW: DeAngelis. PL: Gardner. RH; O'Neill. RV: Suter. GW: Vaughan. DS. (1982).
Methodology for Environmental Risk Analysis. (ORNL/TM-8167). Oak Ridge, TN: Oak Ridge
National Laboratory.
Beckwith. JB. (1999). Human renal carcinoma — Pathogenesis and biology. In CC Capen; E Dybing; JM
Rice; JD Wilbourn (Eds.), IARC Scientific Publication No 147 (pp. 81-93). Lyon, France:
International Agency for Research on Cancer.
Bekele. TG: Zhao. H; Wang. O. (2021). Tissue distribution and bioaccumulation of organophosphate
esters in wild marine fish from Laizhou Bay, North China: Implications of human exposure via
fish consumption. J Hazard Mater 401: 123410. http://dx.doi.Org/10.1016/i.ihazmat.2020.123410
Bekele. TG: Zhao. H; Wang. O: Chen. J. (2019). Bioaccumulation and trophic transfer of emerging
organophosphate flame retardants in the marine food webs of Laizhou Bay, North China.
Environ Sci Technol 53: 13417-13426. http://dx.doi.org/10.1021/acs.est.9b03687
Page 403 of 638
-------
Benotti. MJ; Trenholm. RA; Vanderford. BJ; Hoi adv. JC; Stanford. BP; Snyder. SA. (2009).
Pharmaceuticals and endocrine disrupting compounds in U.S. drinking water. Environ Sci
Technol 43: 597-603. http://dx.doi.org/10.1021/es8Q1845a
Beth-Hubner. M. (1999). Toxicological evaluation and classification of the genotoxic, carcinogenic,
reprotoxic and sensitising potential of tris(2-chloroethyl)phosphate. Int Arch Occup Environ
Health 72: M17-M23. http://dx.doi.org/10.1007/PL00014210
Beyer. A: Mackav. D; Matthies. M; Wania. F; Webster. E. (2000). Assessing long-range transport
potential of persistent organic pollutants. Environ Sci Technol 34: 699-703.
BIBRA. (1977). Microbial mutagenicity test with trichloroethyl phosphate and trichloropropyl
phosphate. (314/1/77). Flushing, NY: Aceto Corporation.
BJB Enterprises. (2017). Safety data sheet: TC-800 part A. Tustin, CA: BJB Enterprises, Inc.
Blessinger. TP: Euling. SY; Wang. L; Hogan. KA; Cai. C: Klinefelter. G: Saillenfait. AM. (2020).
Ordinal dose-response modeling approach for the phthalate syndrome. Environ Int 134: 105287.
http://dx.doi.Org/10.1016/i.envint.2019.105287
Blum. A: Behl. M; Birnbaum. LS: Diamond. ML: Phillips. A: Singla. V: Sipes. NS: Stapleton. HM;
Venier. M. (2019). Organophosphate ester flame retardants: Are they a regrettable substitution
for polybrominated diphenyl ethers? Environ Sci Technol Lett 6: 638-649.
http://dx.doi.org/10.1021/acs.estlett.9b00582
Boeing. (2023). Meeting with Boeing on Risk Management under TSCA Section 6 for
Trichloroethylene, Perchloroethylene, and Tris(2-chloroethyl) Phosphate. (EPA-HQ-OPPT-
2018-0476-0058). Washington, DC: U.S. Environmental Protection Agency.
https://www.regulations.gov/comment/EPA-HQ-OPPT-2018-0476-0Q58
Bonifas. J: Blomeke. B. (2015). N-acetylation of aromatic amines: implication for skin and immune
cells [Review], Frontiers in Bioscience (Elite Edition) 7: 267-281.
http://dx.doi.org/10.2741/E733
Borga. K; Kidd. KA: Muir. DCG: Berglund. O: Conder. JM; Gobas. FAPC: Kucklick. J: Malm. O;
Powell. DE. (2012). Trophic magnification factors: Considerations of ecology, ecosystems, and
study design. Integr Environ Assess Manag 8: 64-84. http://dx.doi.org/10.1002/ieam.244
Bradman. A: Castorina. R; Gaspar. F; Nishioka. M; Colon. M; Weathers. W: Egeghy. PP; Maddalena.
R; Williams. J: Jenkins. PL: McKone. TE. (2014). Flame retardant exposures in California early
childhood education environments. Chemosphere 116: 61-66.
http: //dx. doi. or g/10.1016/i. chemosphere .2014.02.072
Bradman. A: Gaspar. F; Castorina. R; Tong-Lin. E; McKone. T; Maddalena. R. (2012). Environmental
Exposures in Early Childhood Education Environments (pp. 423). (08-305). Berkeley, CA:
University of California, Berkeley, https://www.arb.ca.gov/research/apr/past/08-305.pdf
Brandsma. SH; Leonards. P; Leslie. HA: de Boer. J. (2015). Tracing organophosphorus and brominated
flame retardants and plasticizers in an estuarine food web. Sci Total Environ 505: 22-31.
http://dx.doi.Org/10.1016/i.scitotenv.2014.08.072
Broadwater. K; Ceballos. D; Page. E; Croteau. G: Mueller. C. (2017). Health hazard evaluation report:
HHE-2014-0131-3298, evaluation of occupational exposure to flame retardants at four
gymnastics studios. Broadwater, K; Ceballos, D; Page, E; Croteau, G; Mueller, C.
https://www.cdc. gov/niosh/hhe/reports/pdfs/2014-0131 -3268 .pdf
Brodsky. J: Brodesser. J: Bauer. C: Roembke. J. (1997). The environmental fate of six existing
chemicals in laboratory tests. Chemosphere 34: 515-538. http://dx.doi.org/10.1016/S0Q45-
6535(96)00390-6
Bukowski. K; Wvsokinski. D; Mokra. K; Wozniak. K. (2019). DNA damage and methylation induced
by organophosphate flame retardants: Tris(2-chloroethyl) phosphate and tris(l-chloro-2-propyl)
phosphate in human peripheral blood mononuclear cells. Hum Exp Toxicol 38: 724-733.
http://dx.doi.org/10.1177/0960327119839174
Page 404 of 638
-------
Burka. LT; Sanders. JM; Hern DW; Matthews. HB. (1991). Metabolism of tris(2-chloroethyl) phosphate
in rats and mice. Drug Metab Dispos 19: 443-447.
Buszka. PM; Yeskis. DJ; Kolpin. DW; Furlong. ET; Zaugg. SD; Meyer. MT. (2009). Waste-indicator
and pharmaceutical compounds in landfill-leachate-affected ground water near Elkhart, Indiana,
2000-2002. Bull Environ Contam Toxicol 82: 653-659. http://dx.doi.org/10.1007/sQ0128-009-
9702-z
Caldwell. DJ. (1999). Review of mononuclear cell leukemia in F-344 rat bioassays and its significance
to human cancer risk: A case study using alkyl phthalates [Review], Regul Toxicol Pharmacol
30: 45-53. http://dx.doi.org/10.1006/rtph.1999.1305
Caldwell. DJ: Eldridge. SR; Lington. AW: McKee. RH. (1999). Retrospective evaluation of alpha 2u-
globulin accumulation in male rat kidneys following high doses of diisononyl phthalate. Toxicol
Sci 51: 153-160. http://dx.doi.org/10.1093/toxsci/51.1.153
Canbaz. D; van Velzen. MJ; Hallner. E; Zwinderman. AH: Wickman. M; Leonards. PE; van Ree. R; van
Riit. LS. (2015). Exposure to organophosphate and polybrominated diphenyl ether flame
retardants via indoor dust and childhood asthma. Indoor Air 26: 403-413.
http://dx.doi.Org/10.l 111/ina. 12221
Cao. D; Guo. J: Wang. Y; Li. Z; Liang. K; Corcoran. MB: Hosseini. S: Bonina. SM; Rockne. KJ;
Sturchio. NC: Giesv. JP; Liu. J: Li. A: Jiang. G. (2017). Organophosphate Esters in Sediment of
the Great Lakes. Environ Sci Technol 51: 1441-1449. http://dx.doi.org/10.1021/acs.est.6b05484
Cao. S: Zeng. X: Song. Ha: Li. H; Yu. Z; Sheng. G: Fu. J. (2012). Levels and distributions of
organophosphate flame retardants and plasticizers in sediment from Taihu Lake, China. Environ
Toxicol Chem 31: 1478-1484. http://dx.doi.org/10.1002/etc.1872
Cao. ZG: Zhao. LC: Zhang. YC: Ren. MH: Zhang. YJ: Liu. XT: Jie. JY: Wang. ZY: Li. CH: Shen. MH:
Bu. QW. (2019). Influence of Air Pollution on Inhalation and Dermal Exposure of Human to
Organophosphate Flame Retardants: A Case Study During a Prolonged Haze Episode. Environ
Sci Technol 53: 3880-3887. http://dx.doi.org/10.1021/acs.est.8b07053
Castorina. R; Butt. C: Stapleton. HM; Avery. D; Harlev. KG: Holland. N: Eskenazi. B; Bradman. A.
(2017). Flame retardants and their metabolites in the homes and urine of pregnant women
residing in California (the CHAMACOS cohort). Chemosphere 179: 159-166.
http://dx.doi.Org/10.1016/i.chemosphere.2017.03.076
CDC. (2020). CDC Health Topics A-Z: Healthy food environments: Improving access to healthier food
[Website], https://www.cdc.gov/nutrition/healthv-food-environments/improving-access-to-
healthi er-food. html
CDC. (2021). CDC Health Topics A-Z: Micronutrients [Website],
http s ://www. cdc. gov/nutriti on/mi cronutri ent-
malnutrition/index.html?CDC AA refVal=https%3A%2F%2Fwww.cdc.gov%2Fimmpact%2Fin
dex.html
CDC. (2022). CDC Health Topics A-Z: Physical activity [Website],
http s ://www. cdc. gov/phy si cal activity/index. html
CDC. (2023a). CDC Health Topics A-Z: Cancer [Website], https://www.cdc.gov/cancer/
CDC. (2023b). CDC Health Topics A-Z: Infertility FAQs [Website],
https://www.cdc.gov/reproductivehealth/infertilitv/index.htm
CDC. (2023c). CDC Health Topics A-Z: Nutrition [Website], https://www.cdc.gov/nutrition/index.html
CDC. (2023d). CDC Health Topics A-Z: Stress at work [Website],
https://www.cdc.gov/niosh/topics/stress/
CHAP. (2014). Chronic Hazard Advisory Panel on phthalates and phthalate alternatives (with
appendices). Bethesda, MD: U.S. Consumer Product Safety Commission, Directorate for Health
Sciences. https://www.cpsc.gov/s3fs-public/CHAP-REPORT-With-Appendices.pdf
Page 405 of 638
-------
Chapman. DE; Michener. SR; Powis. G. (1991). Metabolism of the flame retardant plasticizer tris(2-
chloroethyl)phosphate by human and rat liver preparations. Fundam Appl Toxicol 17: 215-224.
http://dx.doi.Org/10.1093/toxsci/17.2.215
Chen. G: Jin. Y; Wu, Y; Liu. L; Fu. Z. (2015a). Exposure of male mice to two kinds of organophosphate
flame retardants (OPFRs) induced oxidative stress and endocrine disruption. Environ Toxicol
Pharmacol 40: 310-318. http://dx.doi.Org/10.1016/i.etap.2015.06.021
Chen. G: Zhang. S: Jin. Y; Wu. Y; Liu. L; Qian. Ft; Fu. Z. (2015b). TPP and TCEP induce oxidative
stress and alter steroidogenesis in TM3 Leydig cells. Reprod Toxicol 57: 100-110.
http://dx.doi.Org/10.1016/i.reprotox.2015.05.011
Chen. S: Tan. Z; Qi. Y; Ouvang. C. (2019a). Sorption of tri-n-butyl phosphate and tris(2-chloroethyl)
phosphate on polyethylene and polyvinyl chloride microplastics in seawater. Mar Pollut Bull
149: 110490. http://dx.doi.Org/10.1016/i.marpolbul.2019.l 10490
Chen. Y; Ye. J: Chen. Ya: Hu. Ha: Zhang. Ft; Ou. H. (2019b). Degradation kinetics, mechanism and
toxicology of tris(2-chloroethyl) phosphate with 185 nm vacuum ultraviolet. Chem Eng J 356:
98-106. http://dx.doi.Org/10.1016/i.cei.2018.09.007
Cheng. W; Sun. L; Huang. W; Ruan. T; Xie. Z; Zhang. P; Ding. R; Li. M. (2013a). Detection and
distribution of Tris(2-chloroethyl) phosphate on the East Antarctic ice sheet. Chemosphere 92:
1017-1021. http://dx.doi.org/10.1016/i.chemosphere.2013.03.038
Cheng. W: Xie. Z; Blais. JM; Zhang. P; Li. M; Yang. C: Huang. W: Ding. R; Sun. L. (2013b).
Organophosphorus esters in the oceans and possible relation with ocean gyres. Environ Pollut
180: 159-164. http://dx.doi.Org/10.1016/i.envpol.2013.05.032
Chokwe. TB; Okonkwo. JO. (2019). Occurrence, distribution and ecological risk assessment of
organophosphorus flame retardants and plasticizers in sediment samples along the Vaal River
catchment, South Africa. Emerging Contaminants 5: 173-178.
http://dx.doi.Org/10.1016/i.emcon.2019.05.003
Choo. G: Cho. HS: Park. K; Lee. JW; Kim. P; Oh. JE. (2018). Tissue-specific distribution and
bioaccumulation potential of organophosphate flame retardants in crucian carp. Environ Pollut
239: 161-168. http://dx.doi.Org/10.1016/i.envpol.2018.03.104
Choo. G: Oh. JE. (2020). Seasonal occurrence and removal of organophosphate esters in conventional
and advanced drinking water treatment plants. Water Res 186: 116359.
http ://dx. doi. org/10.1016/i. watres .2020.116359
Confidential. (1973). Toxicology laboratory report - tris (2-chloroethyl) phosphate [TSCA Submission],
(OTS0516689. 86-890000118S. TSCATS/401616).
https://ntrl.ntis.gov/NTRL/dashboard/searchResults/titleDetail/OTSQ516689.xhtml
Crawford. KA; Hawlev. N: Calafat. AM: Javatilaka. NK; Froehlich. RJ; Has. P; Gallagher. LG: Savitz.
DA: Braun. JM: Werner. EF; Romano. ME. (2020). Maternal urinary concentrations of
organophosphate ester metabolites: associations with gestational weight gain, early life
anthropometry, and infant eating behaviors among mothers-infant pairs in Rhode Island. Environ
Health 19: 97. http://dx.doi.org/10.1186/sl2940-020-0Q648-0
Cristate. J: Alvarez-Martin. A: Rodriguez-Cruz. S: Sanchez-Martin. MJ; Lacorte. S. (2017). Sorption
and desorption of organophosphate esters with different hydrophobicity by soils. Environ Sci
Pollut Res Int 24: 27870-27878. http://dx.doi.org/10.1007/sll356-017-036Q-0
Cristate. J: Aragao Bete. TG: Lacorte. S: de Marchi. MRR. (2019). Occurrence of flame retardants in
landfills: A case study in Brazil. Environ Res 168: 420-427.
http://dx.doi.Org/10.1016/i.envres.2018.10.010
Cristate. J: Garcia Vazquez. A: Barata. C: Lacorte. S. (2013). Priority and emerging flame retardants in
rivers: occurrence in water and sediment, Daphnia magna toxicity and risk assessment. Environ
Int 59: 232-243. http://dx.doi.Org/10.1016/i.envint.2013.06.011
Page 406 of 638
-------
Cristale- J; Ramos. DP; Dantas. RF; Machulek Junior. A; Lacorte. S; Sans. C; Esplugas. S. (2016). Can
activated sludge treatments and advanced oxidation processes remove organophosphorus flame
retardants? Environ Res 144: 11-18. http://dx.doi.Org/10.1016/i.envres.2015.10.008
Delumyea. R; Petel. R. (1979). Deposition velocity of phosphorus-containing particles over southern
Lake Huron, April-October, 1975. Atmos Environ 13: 287-294. http://dx.doi.org/10.1016/00Q4-
6981(79)90171-9
Deng. M: Kuo. DTF: Wu. O: Zhang. Y: Liu. X: Liu. S: Hu. X: Mai. B: Liu. Z: Zhang. H. (2018a).
Organophosphorus flame retardants and heavy metals in municipal landfill leachate treatment
system in Guangzhou, China. Environ Pollut 236: 137-145.
http://dx.doi.Org/10.1016/i.envpol.2018.01.042
Deng. WJ: Li. N: Wu. R; Richard. WKS: Wang. Z; Ho. W. (2018b). Phosphorus flame retardants and
Bisphenol A in indoor dust and PM2.5 in kindergartens and primary schools in Hong Kong.
Environ Pollut 235: 365-371. http://dx.doi.Org/10.1016/i.envpol.2017.12.093
Descartes. D. (2020). Bills of Lading (USA Bills Import HOUSES), http://www.datamyne.com
Dobry. A: Keller. R. (1957). Vapor pressures of some phosphate and phosphonate esters. J Phys Chem
61: 1448-1449. http://dx.doi.org/10.1021/il50556a052
Dodson. RE: Perovich. LJ; Covaci. A: Van den Eede. N: Ionas. AC: Dirtu. AC: Brodv. JG: Rudel. RA.
(2012). After the PBDE phase-out: A broad suite of flame retardants in repeat house dust
samples from California. Environ Sci Technol 46: 13056-13066.
http://dx.doi.org/10.1021/es303879n
Dodson. RE: Udesky. JO: Colton. MP: McCaulev. M; Camann. DE; Yau. AY: Adamkiewicz. G: Rudel.
RA. (2017). Chemical exposures in recently renovated low-income housing: Influence of
building materials and occupant activities. Environ Int 109: 114-127.
http://dx.doi.Org/10.1016/i.envint.2017.07.007
Dodson. RE: Van den Eede. N: Covaci. A: Perovich. LJ: Brodv. JG: Rudel. RA. (2014). Urinary
Biomonitoring of Phosphate Flame Retardants: Levels in California Adults and
Recommendations for Future Studies. Environ Sci Technol 48: 13625-13633.
http://dx.doi.org/10.1021/es503445c
DOE. (2016). Table 1: Chemicals of concern and associated chemical information. PACs. Washington,
DC.
Duncan. M. (2000). Fish consumption survey of the Suquamish Indian Tribe of the Port Madison Indian
Reservation, Puget Sound Region. Suquamish, WA: The Suquamish Tribe, Port Madison Indian
Reservation, http://www.deq.state.or.us/wq/standards/docs/toxics/suquamish2000report.pdf
Duratec. (2018). Safety Data Sheet: Grey fire-resistant primer. Bloomington, CA: Dura Technologies,
Inc.
Dvbing. E; Sanner. T. (1999). Species differences in chemical carcinogenesis of the thyroid gland,
kidney and urinary bladder. In CC Capen; E Dybing; RM Rice; JD Wilbourn (Eds.), IARC
Scientific Publications, vol 147 (pp. 15-32). Lyon, France: International Agency for Research on
Cancer. https://publications.iarc.fr/Book-And-Report-Series/Iarc-Scientific-Publications/Species-
Differences-In-Thvroid-Kidnev-And-Urinarv-Bladder-Carcinogenesis-1999
EC. (2009). Screening Assessment for the Challenge: Ethanol, 2-chloro-, phosphate (3:1) (Tris(2-
chloroethyl) phosphate [TCEP]). http://www.ec.gc.ca/ese-
ees/default.asp?lang=En&xml=C3 78778A-D834-54E0-7F69-E6E2944A74FC
EC/HC. (2011). Screening assessment report on hexabromocyclododecane. Ottawa, Canada.
ECB. (2000). IUCLID dataset: CAS No. 115-96-8: Tris (2-chloroethyl) phosphate. Ispra, Italy:
European Chemicals Bureau, European Commission.
ECB. (2003). Technical guidance document on risk assessment: Part II. (EUR 20418 EN/2).
Luxembourg: Office for Official Publications of the European Communities.
Page 407 of 638
-------
http://ihcp.irc.ec.europa.eu/our activities/public-
health/risk assessment of Biocides/doc/tgd/tgdpart2 2ed.pdf
ECB. (2009). European Union risk assessment report: Tris(2-chloroethyl) phosphate, TCEP (pp. 213).
Italy: European Commission, Office for Official Publications of the European Communities.
https://echa.europa.eu/documents/10162/2663989d-1795-44al-8f50-153a81133258
ECHA. (2008). Guidance on information requirements and chemicals safety assessment: Supporting
reference guidelines, chapters R7-R11. Helsinki, Finland.
http://echa.europa.eu/web/guest/guidance-documents/guidance-on-information-requirements-
and-chemical-safetv-assessment
ECHA. (2018). Screening report: An assessment of whether the use of TCEP, TCPP and TDCP in
articles should be restricted. Helsinki, Finland: European Union, European Chemicals Agency.
https://echa.europa.eu/documents/10162/13641/screening report tcep tcpp td-
cp en.pdf/e0960aa7-f703-499c-24ff-fba627060698
Ekpe. OD; Choo. G: Barcelo. D; Oh. JE. (2020). Chapter One - Introduction of emerging halogenated
flame retardants in the environment. In JE Oh (Ed.), (pp. 1-39). Amsterdam, The Netherlands:
Elsevier, http://dx.doi.org/10.1016/bs.coac.2019.ll.002
EPA Region 9. (2023). Biological evaluation and essential fish habitat assessment for discharges to the
Pacific Ocean from outfalls associated with the City of Los Angeles Hyperion Water
Reclamation Plant and wastewater collection systems. San Francisco, CA.
https://repositorv.librarv.noaa.gov/view/noaa/49053
Fang. M; Webster. TF; Gooden. D; Cooper. EM: McClean. MP: Carignan. C: Makev. C: Stapleton.
HM. (2013). Investigating a novel flame retardant known as V6: measurements in baby products,
house dust, and car dust. Environ Sci Technol 47: 4449-4454.
http://dx.doi.org/10.1021/es400Q32v
FCC. (2021). Meeting with Flame Control Coatings and EPA to discuss use of TCEP in paints and
coatings. (EPA-HQ-OPPT-2018-0476-0056). Washington, DC: U.S. Environmental Protection
Agency. https://www.regulations.gov/comment/EPA-HO-OPPT-2018-0476-0Q56
FDRL. (1972). Cholinesterase studies on rats and rabbits with Olin's intermediate for Chemical 58981
[TSCA Submission], (1-11286-F. OTS0517372. 86-890000046. TSCATS/310359). Olin
Chemicals. https://ntrl.ntis.gov/NTRL/dashboard/searchResults/titleDetail/OTSQ517372.xhtml
Feng. G: Jia. R; Sun. S: Wang. M; Zhao. 0: Liu. L. (2023). Occurrence and treatment effect assessment
of organophosphorus flame retardants in source and drinking water, Jinan, China. Environ Sci
Pollut Res Int 30: 52830-52840. http://dx.doi.org/10.1007/sll356-023-25916-7
Fernie. KJ; Palace. V: Peters. LE; Basu. N: Letcher. RJ; Karouna-Renier. NK; Schultz. SL; Lazarus. RS:
Rattner. BA. (2015). Investigating endocrine and physiological parameters of captive American
kestrels exposed by diet to selected organophosphate flame retardants. Environ Sci Technol 49:
7448-7455. http://dx.doi.org/10.1021/acs.est.5b00857
Follmann. W: Wober. J. (2006). Investigation of cytotoxic, genotoxic, mutagenic, and estrogenic effects
of the flame retardants tris-(2-chloroethyl)-phosphate (TCEP) and tris-(2-chloropropyl)-
phosphate (TCPP) in vitro. Toxicol Lett 161: 124-134.
http://dx.doi.Org/10.1016/i.toxlet.2005.08.008
Foster. SA; Kile. ML: Hystad. P; Diamond. ML: Jantunen. LM; Mandhane. PJ: Moraes. TJ; Navaranian.
G: Scott. JA; Simons. E; Subbarao. P; Takaro. TK; Turvev. SE; Brook. JR. (2024).
Organophosphate ester flame retardants and plasticizers in house dust and mental health
outcomes among Canadian mothers: A nested prospective cohort study in CHILD. Environ Res
240: 117451. http://dx.doi.org/10.1016/i.envres.2023.117451
Fries. E; Mihailovic. I. (2011). Pollution of soils with organophosphorus flame retardants and
plasticizers. J Environ Monit 13: 2692-2694. http://dx.doi.org/10.1039/cleml0538h
Page 408 of 638
-------
Fries. E; Puttmann. W. (2001). Occurrence of organophosphate esters in surface water and ground water
in Germany. J Environ Monit 3: 621-626. http://dx.doi.org/10.1039/bl05072a
Fries. E; Puttmann. W. (2003). Monitoring of the three organophosphate esters TBP, TCEP and TBEP in
river water and ground water (Oder, Germany). J Environ Monit 5: 346-352.
http://dx.doi. org/ 10.103 9/b210342g
Fu. J: Fu. K; Gao. K; Li. H: Xue. O: Chen. Y; Wang. L; Shi. J: Fu. J: Zhang. O: Zhang. A: Jiang. G.
(2020). Occurrence and trophic magnification of organophosphate esters in an Antarctic
ecosystem: Insights into the shift from legacy to emerging pollutants. J Hazard Mater 396:
122742. http://dx.doi.Org/10.1016/i.ihazmat.2020.122742
Galloway. SM; Armstrong. MJ; Reuben. C: Colman. S: Brown. B; Cannon. C: Bloom. AD: Nakamura.
F; Ahmed. M; Duk. S: Rimpo. J: Margolin. BH: Resnick. MA: Anderson. B; Zeiger. E. (1987).
Chromosome aberrations and sister chromatid exchanges in Chinese hamster ovary cells
evaluations of 108 chemicals [Review], Environ Mol Mutagen 10: 1-175.
http://dx.doi.org/10.1002/em.28501005Q2
Gao. Z; Den. Y; Yuan. W: He. Ft; Yang. S: Sun. C. (2014). Determination of organophosphorus flame
retardants in fish by pressurized liquid extraction using aqueous solutions and solid-phase
microextraction coupled with gas chromatography-flame photometric detector. J Chromatogr A
1366: 31-37. http://dx.doi.Org/10.1016/i.chroma.2014.09.028
Giannini. EG: Testa. R; Savarino. V. (2005). Liver enzyme alteration: A guide for clinicians [Review],
CMAJ 172: 367-379. http://dx.doi.org/10.1503/cmai.1040752
Ginsberg. G: Hattis. D; Miller. R; Sonawane. B. (2004). Pediatric pharmacokinetic data: Implications
for environmental risk assessment for children. Pediatrics 113: 973-983.
Giorgino. MJ: Rasmussen. RB; Pfeifle. CM. (2007). Occurrence of organic wastewater compounds in
selected surface-water supplies, Triangle Area of North Carolina, 2002-2005 (pp. 29). (Scientific
Investigations Report 2007-5054). U.S. Geological Survey.
http://dx.doi.org/10.3133/sir20075054
Grimes. G: Beaucham. C: Grant. M; Ramsey. J. (2019). Health hazard evaluation report: HHE-2016-
0257-3333, May 2019, evaluation of exposure to metals and flame retardants at an electronics
recycling company. Grimes, G; Beaucham, C; Grant, M; Ramsey, J.
https://www.cdc.gov/niosh/hhe/reports/pdfs/2016-0257-3333.pdf
Gulati. DK; Russell. S. (1985). Tris(2-chloroethyl) phosphate: Sperm morphology vaginal cytology
evaluations in rodents. (Contract No. N01-ES-3-5026, Study No. SMVCE-82-001).
Environmental Health Research and Testing, Inc.
Guo. J: Romanak. K; Westenbroek. S: Hites. RA; Venier. M. (2017a). Current-use flame retardants in
the water of Lake Michigan tributaries. Environ Sci Technol 51: 9960-9969.
http://dx.doi.org/10.1021/acs.est.7b01294
Guo. J: Venier. M; Salamova. A: Hites. RA. (2017b). Bioaccumulation of Dechloranes,
organophosphate esters, and other flame retardants in Great Lakes fish. Sci Total Environ 583: 1-
9. http://dx.doi.org/10.1016/i.scitotenv.2016.11.063
Guo. JH; Simon. K; Romanak. K; Bowerman. W: Venier. M. (2018). Accumulation of flame retardants
in paired eggs and plasma of bald eagles. Environ Pollut 237: 499-507.
http://dx.doi.Org/10.1016/i.envpol.2018.02.056
Hahladakis. JN; Velis. CA; Weber. R; Iacovidou. E; Purnell. P. (2018). An overview of chemical
additives present in plastics: Migration, release, fate and environmental impact during their use,
disposal and recycling [Review], J Hazard Mater 344: 179-199.
http://dx.doi.Org/10.1016/i.ihazmat.2017.10.014
Hallanger. IG: Sagerup. K; Evenset. A: Kovacs. KM: Leonards. P; Fuglei. E; Routti. H; Aars. J: Strom.
H; Lydersen. C: Gabrielsen. GW. (2015). Organophosphorous flame retardants in biota from
Page 409 of 638
-------
Svalbard, Norway. Mar Pollut Bull 101: 442-447.
http://dx.doi.Org/10.1016/i.marpolbul.2015.09.049
Han. Xu: Hao. Y; Li. Y; Yang. R; Wang. Pu: Zhang. G: Zhang. 0; Jiang. G. (2020). Occurrence and
distribution of organophosphate esters in the air and soils of Ny-Alesund and London Island,
Svalbard, Arctic. Environ Pollut 263: 114495. http://dx.doi.org/10.1016/i.envpol.2020.114495
Hard. GC. (2018). Mechanisms of Rodent Renal Carcinogenesis Revisited [Review], Toxicol Pathol 46:
956-969. http://dx.doi.org/10.1177/0192623318797071
Harper. BL; Harding. AK; Waterhous. T; Harris. SG. (2007). Traditional tribal subsistence exposure
scenario and risk assessment guidance manual. (Grant Number EPA-STAR-J1-R831046).
Corvallis, OR: Oregon State University Department of Public Health and Confederated Tribes of
the Umatilla Indian Reservation.
https://health.oregonstate.edu/sites/health.oregonstate.edu/files/research/pdf/tribal-
grant/exposure scenario and risk guidance manual v2.pdf
Harris. S: Harper. B. (2011). A method for tribal environmental justice analysis. Environmental Justice
4: 231-237. http://dx.doi.org/10.1089/env.2010.0Q38
Haumann. T; Thumulla. J. (2002). Semi volatile organochemicals in indoor environment - Chlorinated
phosphorus and organotin compounds in material und house dust samples [Paper], In 9th
International Conference on Indoor Air Quality and Climate (Indoor Air 2002). Retrieved from
https://www.irbnet.de/daten/iconda/CIB7589.pdf
Haworth. S: Lawlor. T; Mortelmans. K; Speck. W: Zeiger. E. (1983). Salmonella mutagenicity test
results for 250 chemicals. Environ Mutagen 5: 3-142. http://dx.doi.org/10.1002/em.28600507Q3
Havnes. WM. (2014). Tris(2-chloroethyl) phosphate. In WMHaynes (Ed.), CRC handbook of chemistry
and physics (95th ed., pp. 3-542). Boca Raton, FL: CRC Press.
Hazleton Laboratories. (1983). Screening of priority chemicals for potential reproductive hazard (final
report) with attachments and cover sheet [TSCA Submission], (Hazleton Study No. 6125-101
through 6125-110. OTS0516205. 86-870001624. TSCATS/400290). Shell Oil Company.
https://ntrl.ntis. gov/NTRL/dashboard/searchResults.xhtml?searchOuerv=OTSQ516205
He. C: Covaci. A: Heffernan. AL; Baduel. C: Harden. FA: Mueller. JF; Toms. LML; Nele Van Den. E;
Hobson. P; Thai. P; Wang. X: Li. Y. (2018a). Urinary metabolites of organophosphate esters:
Concentrations and age trends in Australian children. Environ Int 111: 124-130.
http://dx.doi.Org/10.1016/i.envint.2017.l 1.019
He. C: Wang. X: Tang. S: Phong. T; Li. Z; Baduel. C: Mueller. JF. (2018b). Concentrations of
Organophosphate Esters and Their Specific Metabolites in Food in Southeast Queensland,
Australia: Is Dietary Exposure an Important Pathway of Organophosphate Esters and Their
Metabolites? Environ Sci Technol 52: 12765-12773. http://dx.doi.org/10.1021 /acs.est.8b03043
He. MJ: Yang. T; Yang. ZH; Li. O: Wei. SO. (2017). Occurrence and Distribution of Organophosphate
Esters in Surface Soil and Street Dust from Chongqing, China: Implications for Human
Exposure. Arch Environ Contam Toxicol 73: 349-361. http://dx.doi.org/10.1007/sQ0244-017-
0432-7
He. RW: Li. YZ; Xiang. P; Li. C: Cui. XY; Ma. LQ. (2018c). Impact of particle size on distribution and
human exposure of flame retardants in indoor dust. Environ Res 162: 166-172.
http://dx.doi.Org/10.1016/i.envres.2017.12.014
Health Canada. (2020). Screening assessment - Phthalate substance grouping. (Enl4-393/2019E-PDF).
Environment and Climate Change Canada, https://www.canada.ca/en/environment-climate-
change/services/evaluating-existing-substances/screening-assessment-phthalate-substance-
grouping.html
Heindel. JJ; Lamb. JC: Chapin. RE: Gulati. DK; Hope. E; George. J: Jameson. CY; Teague. J: Schwetz.
BA. (1989). Reproductive toxicity testing by continuous breeding test protocol in Swiss (CD-I)
Page 410 of 638
-------
mice. (PB89152425; NTP-89054). Springfield, VA: U.S. Department of Commerce.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults/titleDetail/PB89152425.xhtml
Hernandez-Castro. I; Eckel. SP; Howe. CG; Niu. Z; Kannan. K; Robinson. M; Foley. HB; Yang. T;
Vigil. MJ; Chen. X; Grubbs. B; Lerner. D; Lurvev. N; Al-Maravati. L; Habre. R; Dunton. GF;
Farzan. SF; Aung. MT; Breton. CV; Bastain. TM. (2023a). Prenatal exposures to
organophosphate ester metabolite mixtures and children's neurobehavioral outcomes in the
MADRES pregnancy cohort. Environ Health 22: 66. http://dx.doi.org/ 10.1186/sl2940-023-
01017-3
Hernandez-Castro. I; Eckel. SP: Howe. CG: Niu. ZZ; Kannan. K; Robinson. M; Foley. HB: Grubbs. B;
Al-Maravati. L: Lerner. D: Lurvev. N: Aung. MT: Habre. R: Dunton. GF: Farzan. SF: Breton.
CV: Bastain. TM. (2023b). Sex-specific effects of prenatal organophosphate ester (OPE)
metabolite mixtures and adverse infant birth outcomes in the maternal and developmental risks
from environmental and social stressors (MADRES) pregnancy cohort. Environ Res 226:
115703. http://dx.doi.Org/10.1016/i.envres.2023.l 15703
Herr. DW: Sanders. JM: Matthews. HB. (1991). Brain distribution and fate of tris(2-chloroethyl)
phosphate in Fischer 344 rats. Drug Metab Dispos 19: 436-442.
Herrlich. T; Steib. C: Lang. A. (2013). [Flame-resistant coating for the rear side of a carpet] [Patent],
Clariant International Ltd. https://patents.google.com/patent/WO2014Q26741Al/en
Hill. AB. (1965). The environment and disease: Association or causation? Proc R Soc Med 58: 295-300.
http://dx.doi.org/10.1177/0035915765058005Q3
Hoehn. RM: Jahl. LG: Herkert. NJ: Hoffman. K: Soehl. A: Diamond. ML: Blum. A: Stapleton. HM.
(2024). Flame retardant exposure in vehicles is influenced by use in seat foam and temperature.
Environ Sci Technol 58: 8825-8834. http://dx.doi.org/10.1021/acs.est.3cl0440
Hoffman. K: Lorenzo. A: Butt. CM: Hammel. SC: Henderson. BB: Roman. SA: Scheri. RP; Stapleton.
HM: Sosa. JA. (2017). Exposure to flame retardant chemicals and occurrence and severity of
papillary thyroid cancer: A case-control study. Environ Int 107: 235-242.
http://dx.doi.Org/10.1016/i.envint.2017.06.021
Hou. R: Liu. C: Gao. X: Xu. Y; Zha. J: Wang. Z. (2017). Accumulation and distribution of
organophosphate flame retardants (PFRs) and their di-alkyl phosphates (DAPs) metabolites in
different freshwater fish from locations around Beijing, China. Environ Pollut 229: 548-556.
http://dx.doi.Org/10.1016/i.envpol.2017.06.097
HSDB. (2015). Tris(2-chloroethyl) phosphate, CASRN: 115-96-8. U.S. Department of Health and
Human Services, National Institutes of Health, National Library of Medicine.
http://toxnet.nlm.nih. gov/cgi-bin/sis/search2/r?dbs+hsdb:@term+@DOCNO+2577
Hu. F; Li. W: Wang. H: Peng. H: He. J: Ding. J: Zhang. W. (2023). Environmentally relevant
concentrations of tris (2-chloroethyl) phosphate (TCEP) induce hepatotoxicity in zebrafish
(Danio rerio): a whole life-cycle assessment. Fish Physiol Biochem 49: 1421-1433.
http://dx.doi.org/10.1007/slQ695-023-01265-7
Hu. F; Zhao. Y; Dong. F; Wang. H: Zheng. M: Zhang. W: Chen. X. (2022). Insights into the
mechanisms of tris(2-chloroethyl) phosphate-induced growth inhibition in juvenile yellow
catfish Pelteobagrus fulvidraco. Aquat Toxicol 247: 106170.
http ://dx.doi. org/10.1016/i.aquat ox.2022.106170
Hurtado. C: Montano-Chavez. YN: Dominguez. C: Bavona. JM. (2017). Degradation of Emerging
Organic Contaminants in an Agricultural Soil: Decoupling Biotic and Abiotic Processes. Water
Air Soil Pollut 228: 243-p. 243. http://dx.doi.org/10.1007/sll270-017-34Q2-9
Hutchins. SR: Tomson. MB: Wilson. JT: Ward. CH. (1984). Fate of trace organics during rapid
infiltration of primary waste water at Fort Devens, Massachusetts (USA). Water Res 18: 1025-
1036. http://dx.doi.org/10.1016/0043-1354(84)90255-0
Page 411 of 638
-------
I ARC. (1990). Tris(2-chloroethyl) phosphate [IARC Monograph], In IARC Monographs on the
Evaluation of Carcinogenic Risks to Humans, vol 48 (pp. 109-120). Lyon, France.
https://publications.iarc.fr/Book-And-Report-Series/Iarc-Monographs-On-The-Identification-Of-
Carcinogenic-Hazards-To-Humans/Some-Flame-Retardants-And-Textile-Chemicals-And-
Exposures-In-The-Textile-Manufacturing-Industry-1990
IARC. (2019). Report of the Advisory Group to Recommend Priorities for the IARC Monographs
during 2020-2024. Geneva, Switzerland: World Health Organization.
https://www.iarc.who.int/news-events/report-of-the-advisorv-group-to-recommend-priorities-for-
the-iarc-monographs-during-2020-2024/
ILO. (2019). International Chemical Safety Card: 1677. Tris(2-chloroethyl) phosphate.
http://www.ilo.org/dvn/icsc/showcard.displav7p lang=en&p card id=1677&p version=2
Ingerowski. G: Friedle. A: Thumulla. J. (2001). Chlorinated ethyl and isopropyl phosphoric acid triesters
in the indoor environment—an inter-laboratory exposure study. Indoor Air 11: 145-149.
http://dx.doi.Org/10.1034/i.1600-0668.2001.011003145.x
Ionas. AC: Dirtu. AC: Anthonissen. T; Neels. H; Covad. A. (2014). Downsides of the recycling process:
Harmful organic chemicals in children's toys. Environ Int 65: 54-62.
http://dx.doi.Org/10.1016/i.envint.2013.12.019
IPCS. (1998). Flame retardants: Tris(chloropropyl) phosphate and tris(2-chloroethyl) phosphate.
(Environmental Health Criteria 209). Geneva, Switzerland: World Health Organization.
http://www.inchem.org/documents/ehc/ehc/ehc209.htm
Jackson. J: Sutton. R. (2008). Sources of endocrine-disrupting chemicals in urban wastewater, Oakland,
CA. Sci Total Environ 405: 153-160. http://dx.doi.Org/10.1016/i.scitotenv.2008.06.033
Javatilaka. NK; Restrepo. P; Williams. L; Ospina. M; Valentin-Blasini. L; Calafat. AM. (2017).
Quantification of three chlorinated dialkyl phosphates, diphenyl phosphate, 2,3,4,5-
tetrabromobenzoic acid, and four other organophosphates in human urine by solid phase
extraction-high performance liquid chromatography-tandem mass spectrometry. Anal Bioanal
Chem 409: 1323-1332. http://dx.doi.org/10.1007/s00216-016-0Q61-4
Ji. Y: Wang. Y: Yao. Y: Ren. C: Lan. Z: Fang. X: Zhang. K: Sun. W: Alder. AC: Sun. H. (2019).
Occurrence of organophosphate flame retardants in farmland soils from Northern China: Primary
source analysis and risk assessment. Environ Pollut 247: 832-838.
http://dx.doi.Org/10.1016/i.envpol.2019.01.036
Johnson. PD; Goldberg. SJ: Mays. MZ; Dawson. BY. (2003). Threshold of trichloroethylene
contamination in maternal drinking waters affecting fetal heart development in the rat. Environ
Health Perspect 111: 289-292. http://dx.doi.org/10.1289/ehp.5125
Kademoglou. K; Xu. F; Padilla-Sanchez. JA; Haug. LS: Covaci. A: Collins. CD. (2017). Legacy and
alternative flame retardants in Norwegian and UK indoor environment: Implications of human
exposure via dust ingestion. Environ Int 102: 48-56.
http://dx.doi.Org/10.1016/i.envint.2016.12.012
Kaiiwara. N: Noma. Y; Takigami. H. (2011). Brominated and organophosphate flame retardants in
selected consumer products on the Japanese market in 2008. J Hazard Mater 192: 1250-1259.
http://dx.doi.Org/10.1016/i.ihazmat.2011.06.043
Kang. H; Lee. J: Lee. JP; Choi. K. (2019). Urinary metabolites of organophosphate esters (OPEs) are
associated with chronic kidney disease in the general US population, NHANES 2013-2014.
Environ Int 131: 5034-5034. http://dx.doi.Org/10.1016/i.envint.2019.105034
Kawagoshi. Y; Fukunaga. I: Itoh. H. (1999). Distribution of organophosphoric acid triesters between
water and sediment at a sea-based solid waste disposal site. Journal of Material Cycles and
Waste Management 1: 53-61. http://dx.doi.org/10.1007/slQ163-999-0005-6
Page 412 of 638
-------
Kawagoshi. Y; Nakamura. S; Fukunaga. I. (2002). Degradation of organophosphoric esters in leachate
from a sea-based solid waste disposal site. Chemosphere 48: 219-225.
http://dx.doi.org/10.1016/S0045-6535(02X)0051-6
Kawashima. K; Tanaka. S: Nakaura. S: Nagao. S: Endo. T: Onoda. K; Takanaka. A: Omori. Y. (1983).
[Effect of oral administration of tris(2-chloroethyl) phosphate to pregnant rats on prenatal and
postnatal development], Eisei Shikenjo Hokoku 101: 55-61.
Kerler. F; Schoenherr. J. (1988). Permeation of lipophilic chemicals across plant cuticles prediction
from partition coefficients and molar volumes. Arch Environ Contam Toxicol 17: 7-12.
http://dx.doi.org/10.1007/BF01Q55147
Kerric. A: Okeme. J: Jantunen. L; Giroux. JF: Diamond. ML: Verreault. J. (2021). Spatial and temporal
variations of halogenated flame retardants and organophosphate esters in landfill air: Potential
linkages with gull exposure. Environ Pollut 271: 116396.
http://dx.doi. org/10.1016/i. envpol .2020.116396
Kim. J: Isobe. T; Muto. M; Nguyen Minh. T; Katsura. K; Malarvannan. G: Sudaryanto. A: Chang. KH:
Prudente. M; Pham Hung. V: Takahashi. S: Tanabe. S. (2014). Organophosphorus flame
retardants (PFRs) in human breast milk from several Asian countries. Chemosphere 116: 91-97.
http://dx.doi. org/10.1016/i. chemosphere.2014.02.03 3
Kim. SD; Cho. J: Kim. IS: Vanderford. BJ; Snyder. SA. (2007). Occurrence and removal of
pharmaceuticals and endocrine disruptors in South Korean surface, drinking, and waste waters.
Water Res 41: 1013-1021. http://dx.doi.Org/10.1016/i.watres.2006.06.034
Kim. UJ: Kannan. K. (2018). Occurrence and distribution of organophosphate flame
retardants/plasticizers in surface waters, tap water, and rainwater: Implications for human
exposure. Environ Sci Technol 52: 5625-5633. http://dx.doi.org/10.1021/acs.est.8b00727
Kim. UJ: Oh. JK; Kannan. K. (2017). Occurrence, removal, and environmental emission of
organophosphate flame retardants/plasticizers in a wastewater treatment plant in New York
State. Environ Sci Technol 51: 7872-7880. http://dx.doi.org/10.1021/acs.est.7b02035
Kim. Un; Jo. H; Lee. In: Joo. G: Oh. J. (2015). Investigation of bioaccumulation and biotransformation
of polybrominated diphenyl ethers, hydroxylated and methoxylated derivatives in varying
trophic level freshwater fishes. Chemosphere 137: 108-114.
http: //dx. doi. or g/10.1016/i. chemosphere .2015.05.104
Koiadinovic. J: Potier. M; Le Corre. M; Cosson. RP; Bustamante. P. (2007). Bioaccumulation of trace
elements in pelagic fish from the Western Indian Ocean. Environ Pollut 146: 548-566.
http://dx.doi.Org/10.1016/i.envpol.2006.07.015
Kolpin. DW: Furlong. ET; Meyer. MT; Thurman. EM: Zaugg. SD: Barber. LB: Buxton. HT. (2002).
Pharmaceuticals, hormones, and other organic wastewater contaminants in US streams, 1999-
2000: A national reconnaissance. Environ Sci Technol 36: 1202-1211.
http://dx.doi.org/10.1021/es011055i
Krivoshiev. BY: Beemster. GTS: Sprangers. K; Blust. R; Husson. SJ. (2018). A toxicogenomics
approach to screen chlorinated flame retardants tris(2-chloroethyl) phosphate and tris(2-
chloroisopropyl) phosphate for potential health effects. J Appl Toxicol 38: 459-470.
http://dx.doi.org/10.1002/iat.3553
Krivoshiev. BY: Dardenne. F: Covaci. A: Blust. R: Husson. SJ. (2016). Assessing in-vitro estrogenic
effects of currently-used flame retardants. Toxicol In Vitro 33: 153-162.
http://dx.doi.Org/10.1016/i.tiv.2016.03.006
Kurt-Karakus. P: Alegria. H: Birgul. A: Gungormus. E: Jantunen. L. (2018). Organophosphate ester
(OPEs) flame retardants and plasticizers in air and soil from a highly industrialized city in
Turkey. Sci Total Environ 625: 555-565. http://dx.doi.org/10.1016/i.scitotenv.2017.12.307
Page 413 of 638
-------
La Guardia. MJ; Hale. RC. (2015). Halogenated flame-retardant concentrations in settled dust,
respirable and inhalable particulates and polyurethane foam at gymnastic training facilities and
residences. Environ Int 79: 106-114. http://dx.doi.Org/10.1016/i.envint.2015.02.014
Lai. S: Xie. Z; Song. T; Tang. J: Zhang. Y; Mi. W: Peng. J: Zhao. Y; Zou. S: Ebinghaus. R. (2015).
Occurrence and dry deposition of organophosphate esters in atmospheric particles over the
northern South China Sea. Chemosphere 127: 195-200.
http: //dx. doi. or g/10.1016/i. chemosphere .2015.02.015
Lanning. LL; Creasy. DM: Chapin. RE: Mann. PC: Barlow. NJ: Regan. KS: Goodman. DG. (2002).
Recommended approaches for the evaluation of testicular and epididymal toxicity. Toxicol
Pathol 30: 507-520. http://dx.doi.org/10.1080/01926230290105695
Lapworth. DJ; Baran. N: Stuart. ME: Ward. RS. (2012). Emerging organic contaminants in
groundwater: A review of sources, fate and occurrence [Review], Environ Pollut 163: 287-303.
http://dx.doi.Org/10.1016/i.envpol.2011.12.034
Lee. E; Shon. HK; Cho. J. (2014). Role of wetland organic matters as photosensitizer for degradation of
micropollutants and metabolites. J Hazard Mater 276: 1-9.
http://dx.doi.Org/10.1016/i.ihazmat.2014.05.001
Lee. RE. Jr.; Patterson. RK. (1969). Size determination of atmospheric phosphate, nitrate, chloride, and
ammonium particulate in several urban areas. Atmos Environ 3: 249-255.
Lee. S: Cho. HJ; Choi. W: Moon. HB. (2018). Organophosphate flame retardants (OPFRs) in water and
sediment: Occurrence, distribution, and hotspots of contamination of Lake Shihwa, Korea. Mar
Pollut Bull 130: 105-112. http://dx.doi.Org/10.1016/i.marpolbul.2018.03.009
Lehner. AF; Samsing. F; Rumbeiha. WK. (2010). Organophosphate ester flame retardant-induced acute
intoxications in dogs. J Med Toxicol 6: 448-458. http://dx.doi.org/10.10Q7/sl3181-010-0105-7
Lei. YD: Wania. F. (2004). Is rain or snow a more efficient scavenger of organic chemicals? Atmos
Environ 38: 3557-3571. http://dx.doi.Org/10.1016/i.atmosenv.2004.03.039
Lewis. RJ. Sr.; Hawlev. GG. (2007). Hawley's condensed chemical dictionary (15th ed.). Hoboken, NJ:
John Wiley & Sons. http://dx.doi.org/10.1002/978Q470114735
Li. C; Chen. J; Xie. HB; Zhao. Y; Xia. D; Xu. T; Li. X; Qiao. X. (2017a). Effects of Atmospheric Water
on OH-initiated Oxidation of Organophosphate Flame Retardants: A DFT Investigation on
TCPP. Environ Sci Technol 51: 5043-5051. http://dx.doi.org/10.1021/acs.est.7b00347
Li. J; Xie. Z; Mi. W; Lai. S; Tian. C; Emeis. KC; Ebinghaus. R. (2017b). Organophosphate esters in air,
snow, and seawater in the North Atlantic and the arctic. Environ Sci Technol 51: 6887-6896.
http://dx.doi.org/10.1021/acs.est.7b01289
Li. M; Mathieu. J; Yang. Y; Fiorenza. S; Deng. Y; He. Z; Zhou. J; Alvarez. PJ. (2013). Widespread
distribution of soluble di-iron monooxygenase (SDIMO) genes in Arctic groundwater impacted
by 1,4-dioxane. Environ Sci Technol 47: 9950-9958. http://dx.doi.org/ 10.1021/es402228x
Li. M; Van Orden. ET; Devries. DJ; Xiong. Z; Hinchee. R; Alvarez. PJ. (2015). Bench-scale
biodegradation tests to assess natural attenuation potential of 1,4-dioxane at three sites in
California. Biodegradation 26: 39-50. http://dx.doi.org/10.1007/slQ532-014-9714-l
Li. TY; Bao. LJ; Wu. CC; Liu. LY; Wong. CS; Zeng. EY. (2019a). Organophosphate flame retardants
emitted from thermal treatment and open burning of e-waste. J Hazard Mater 367: 390-396.
http://dx.doi.Org/10.1016/i.ihazmat.2018.12.041
Li. W; Wang. Y; Kannan. K. (2019b). Occurrence, distribution and human exposure to 20
organophosphate esters in air, soil, pine needles, river water, and dust samples collected around
an airport in New York state, United States. Environ Int 131: 105054.
http://dx.doi.Org/10.1016/i.envint.2019.105054
Li. Y; Fu. Y; Hu. K; Zhang. Y; Chen. J; Zhang. S; Zhang. B; Liu. Y. (2020). Positive correlation
between human exposure to organophosphate esters and gastrointestinal cancer in patients from
Page 414 of 638
-------
Wuhan, China. Ecotoxicol Environ Saf 196: 110548.
http: //dx. doi. or g/ 10.1016/i. ecoenv .2020.110548
Liang. K; Liu. J. (2016). Understanding the distribution, degradation and fate of organophosphate esters
in an advanced municipal sewage treatment plant based on mass flow and mass balance analysis.
Sci Total Environ 544: 262-270. http: //dx. doi. or g/10.1016/i. scitotenv. 2015.11.112
Liao. K; Zhao. Y; Qu. J: Yu. W: Hu. S: Fang. S: Zhao. M; Jin. H. (2023). Organophosphate esters
concentrations in human serum and their associations with Sjogren syndrome. Environ Pollut
331: 121941. http://dx.doi.Org/10.1016/i.envpol.2023.121941
Life Sciences Research Ltd. (1990a). Fyrol CEF: Acute toxicity to rainbow trout [TSCA Submission],
(90/AKL021/0161. OTS0528355. 86-910000061. TSCATS/413338). Life Science Research Ltd.
Akzo Chemicals Inc.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults/titleDetail/OTSQ528355.xhtml
Life Sciences Research Ltd. (1990b). Fyrol CEF: Assessment of its ready biodegradability. Modified
Sturm test [TSCA Submission], (90/AKL023/0230. OTS0528355. 86-910000061.
TSCATS/413338). Akzo Chemicals Inc.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults/titleDetail/OTSQ528355.xhtml
Life Sciences Research Ltd. (1990c). Fyrol FR-2: Assessment of its ready biodegradability. Modified
Sturm test [TSCA Submission], (90/AKL029/0232. OTS0528355. 86-910000061.
TSCATS/413338). Akzo Chemicals Inc.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults/titleDetail/OTSQ528355.xhtml
Lipworth. BJ. (1996). Pharmacokinetics of inhaled drugs [Review], Br J Clin Pharmacol 42: 697-705.
http://dx.doi.Org/10.1046/i.1365-2125.1996.00493.x
Liu. LY; He. K; Hites. RA; Salamova. A. (2016). Hair and nails as noninvasive biomarkers of human
exposure to brominated and organophosphate flame retardants. Environ Sci Technol 50: 3065-
3073. http://dx.doi.org/10.1021/acs.est.5b05073
Liu. LY: Salamova. A: He. K; Hites. RA. (2015). Analysis of polybrominated diphenyl ethers and
emerging halogenated and organophosphate flame retardants in human hair and nails. J
Chromatogr A 1406: 251-257. http://dx.doi.Org/10.1016/i.chroma.2015.06.003
Liu. M; Li. A: Meng. L; Zhang. G: Guan. X: Zhu. J: Li. Y; Zhang. O: Jiang. G. (2022). Exposure to
novel brominated flame retardants and organophosphate esters and associations with thyroid
cancer risk: A case-control study in eastern China. Environ Sci Technol 56: 17825-17835.
http://dx.doi.org/10.1021/acs.est.2c04759
Liu. Y; Li. Y; Dong. S: Han. L; Guo. R; Fu. Y; Zhang. S: Chen. J. (2021). The risk and impact of
organophosphate esters on the development of female-specific cancers: Comparative analysis of
patients with benign and malignant tumors. J Hazard Mater 404: 124020.
http://dx.doi.Org/10.1016/i.ihazmat.2020.124020
Liu. Y; Liggio. J: Harner. T; Jantunen. L; Shoeib. M; Li. SM. (2014). Heterogeneous OH initiated
oxidation: A possible explanation for the persistence of organophosphate flame retardants in air.
Environ Sci Technol 48: 1041-1048. http://dx.doi.org/10.1021/es404515k
Liu. Y; Luo. X: Corella. PZ; Zeng. Y; Mai. BH. (2019a). Organophosphorus flame retardants in a
typical freshwater food web: Bioaccumulation factors, tissue distribution, and trophic transfer.
Environ Pollut 255 Pt. 2: 113286. http://dx.doi.Org/10.1016/i.envpol.2019.l 13286
Liu. YE: Luo. XJ; Huang. LQ; Zeng. YH; Mai. BX. (2019b). Organophosphorus flame retardants in fish
from Rivers in the Pearl River Delta, South China. Sci Total Environ 663: 125-132.
http: //dx. doi. or g/10.1016/i. scitotenv .2019.01.344
Lown. JW; Joshua. AY: McLaughlin. LW. (1980). Novel antitumor nitrosoureas and related compounds
and their reactions with DNA. J Med Chem 23: 798-805.
http://dx.doi.org/10.1021/im0Q181a018
Page 415 of 638
-------
Luo. P; Bao. LJ; Guo. Y; Li. SM; Zeng. EY. (2015). Size-dependent atmospheric deposition and
inhalation exposure of particle-bound organophosphate flame retardants. J Hazard Mater 301:
504-511. http://dx.doi.Org/10.1016/i.ihazmat.2015.09.014
Ma. J: Zhu. H; Kannan. K. (2019). Organophosphorus flame retardants and plasticizers in breast milk
from the United States. Environ Sci Technol Lett 6: 525-531.
http: //dx. doi. or g/10.1021 /acs. estl ett. 9b003 94
Ma. Y; Voita. S: Becanova. J: Habtemichael. AZ; Adelman. DA: Muir. D; Lohmann. R. (2021). Spatial
distribution and air-water exchange of organophosphate esters in the lower Great Lakes. Environ
Pollut 286: 117349. http://dx.doi.Org/10.1016/i.envpol.2021.117349
Ma. Y; Xie. Z; Lohmann. R; Mi. W: Gao. G. (2017). Organophosphate ester flame retardants and
plasticizers in ocean sediments from the north pacific to the arctic ocean. Environ Sci Technol
51: 3809-3815. http://dx.doi.org/10.1021/acs.est.7b00755
Mackav. D; Di Guardo. A: Paterson. S: Cowan. CE. (1996). Evaluating the environmental fate of a
variety of types of chemicals using the EQC model. Environ Toxicol Chem 15: 1627-1637.
http://dx.doi.org/10.1002/etc.562015Q929
Maddela. NR; Venkateswarlu. K; Megharai. M. (2020). Tris(2-chloroethyl) phosphate, a pervasive
flame retardant: Critical perspective on its emissions into the environment and human toxicity
[Review], Environ Sci Process Impacts 22: 1809-1827. http://dx.doi.org/10.1039/d0em00222d
Mage. DT; Allen. RH; Kodali. A. (2008). Creatinine corrections for estimating children's and adult's
pesticide intake doses in equilibrium with urinary pesticide and creatinine concentrations. J Expo
Sci Environ Epidemiol 18: 360-368. http://dx.doi.org/10.1038/si.ies.7500614
Makinen. MSE; Makinen. MR A; Koistinen. JTB; Pasanen. AL; Pasanen. PO; Kalliokoski. PJ; Korpi.
AM. (2009). Respiratory and dermal exposure to organophosphorus flame retardants and
tetrabromobisphenol A at five work environments. Environ Sci Technol 43: 941-947.
http://dx.doi.org/10.1021/es802593t
Mansouri. K; Grulke. CM: Judson. RS: Williams. AJ. (2018). OPERA models for predicting
physicochemical properties and environmental fate endpoints. J Cheminform 10: 10.
http://dx.doi.org/10.1186/sl3321-018-0263-l
Marklund. A: Anders son. B; Haglund. P. (2003). Screening of organophosphorus compounds and their
distribution in various indoor environments. Chemosphere 53: 1137-1146.
http://dx.doi.org/10.1016/50045-6535(03)00666-0
Marklund. A: Anders son. B; Haglund. P. (2005a). Organophosphorus flame retardants and plasticizers
in Swedish sewage treatment plants. Environ Sci Technol 39: 7423-7429.
http://dx.doi.org/10.1021/es0510131
Marklund. A: Anders son. B; Haglund. P. (2005b). Traffic as a source of organophosphorus flame
retardants and plasticizers in snow. Environ Sci Technol 39: 3555-3562.
http://dx.doi.org/10.1021/esQ482177
Maronpot. RR; Nvska. A: Foreman. JE; Ramot. Y. (2016). The legacy of the F344 rat as a cancer
bioassay model (a retrospective summary of three common F344 rat neoplasms) [Review], Crit
Rev Toxicol 46: 641-675. http://dx.doi.org/10.1080/10408444.2Q16.1174669
Maruva. KA; Dodder. NG: Sengupta. A: Smith. DJ; Lyons. JM; Heil. AT: Drewes. JE. (2016).
Multimedia screening of contaminants of emerging concern (CECS) in coastal urban watersheds
in southern California (USA). Environ Toxicol Chem 35: 1986-1994.
http://dx.doi.org/10.1002/etc.3348
Masoner. JR; Kolpin. DW: Furlong. ET; Cozzarelli. IM; Gray. JL. (2016). Landfill leachate as a mirror
of today's disposable society: Pharmaceuticals and other contaminants of emerging concern in
final leachate from landfills in the conterminous United States. Environ Toxicol Chem 35: 906-
918. http://dx.doi.org/10.1002/etc.3219
Page 416 of 638
-------
Masoner. JR; Kolpin. DW; Furlong. ET; Cozzarelli. IM; Gray. JL; Schwab. EA. (2014a). Contaminants
of emerging concern in fresh leachate from landfills in the conterminous United States. Environ
Sci Process Impacts 16: 2335-2354. http://dx.doi.org/10.1039/c4em00124a
Masoner. JR: Kolpin. DW: Furlong. ET: Cozzarelli. IM: Gray. JL: Schwab. EA. (2014b). Contaminants
of emerging concern in fresh leachate from landfills in the conterminous United States:
Supplementary information [Supplemental Data], Environ Sci Process Impacts 16.
Mato. Y: Isobe. T; Takada. H: Kanehiro. H: Ohtake. C: Kaminuma. T. (2001). Plastic resin pellets as a
transport medium for toxic chemicals in the marine environment. Environ Sci Technol 35: 318-
324. http://dx.doi.org/10.1021/esOO 10498
Matsson. P; Kihlberg. J. (2017). How big is too big for cell permeability? J Med Chem 60: 1662-1664.
http://dx.doi.org/10.1021/acs.imedchem.7b00237
Matsukami. H; Nguyen Minh. T; Suzuki. Go: Someva. M; Le Huu. T; Pham Hung. V: Takahashi. S:
Tanabe. S: Takigami. H. (2015). Flame retardant emission from e-waste recycling operation in
northern Vietnam: Environmental occurrence of emerging organophosphorus esters used as
alternatives for PBDEs. Sci Total Environ 514: 492-499.
http: //dx. doi. or g/10.1016/i. scitotenv .2015.02.008
Matthews. HB; Dixon. D: Herr. DW: Tilson. H. (1990). Subchronic toxicity studies indicate that tris(2-
chloroethyl)phosphate administration results in lesions in the rat hippocampus. Toxicol Ind
Health 6: 1-15. http://dx.doi.org/10.1177/07482337900060Q101
Matthews. HB: Eustis. SL: Haseman. J. (1993). Toxicity and carcinogenicity of chronic exposure to
tris(2-chloroethyl)phosphate. Fundam Appl Toxicol 20: 477-485.
http://dx.doi.org/10.1006/faat.1993.1058
Mayer. AC: Fent. KW: Chen. IC: Sammons. D: Toennis. C: Robertson. S: Kerber. S: Horn. GP: Smith.
PL: Calafat. AM: Ospina. M: Siodin. A. (2021). Characterizing exposures to flame retardants,
dioxins, and furans among firefighters responding to controlled residential fires. Int J Hyg
Environ Health 236: 113782. http://dx.doi.Org/10.1016/i.iiheh.2021.113782
McDonough. CA: De Silva. AO: Sun. C: Cabrerizo. A: Adelman. D: Soltwedel. T; Bauerfeind. E; Muir.
DCG: Lohmann. R. (2018). Dissolved organophosphate esters and polybrominated diphenyl
ethers in remote marine environments: Arctic surface water distributions and net transport
through fram strait. Environ Sci Technol 52: 6208-6216.
http://dx.doi.org/10.1021/acs.est.8b01127
Mehvar. R. (2018). Clearance concepts: Fundamentals and application to pharmacokinetic behavior of
drugs [Review], J Pharm Pharm Sci 21: 88s-102s. http://dx.doi.org/10.18433/ipps29896
Melnikova. PL: Nesmelova. IV: Skirda. YD. (2019). Effect of reducing agent TCEP on translational
diffusion and supramolecular assembly in aqueous solutions of a-casein. J Phys Chem B 123:
2305-2315. http://dx.doi.org/10.1021/acs.ipcb.8bl2511
Mendv. A: Percy. Z: Braun. JM: Lanphear. B: La Guardia. MJ: Hale. RC: Yolton. K: Chen. A. (2024).
Prenatal exposure to replacement flame retardants and organophosphate esters and childhood
adverse respiratory outcomes. Environ Res 240: 117523.
http: //dx. doi. or g/10.1016/i. envres .2023.117523
Mennillo. E: Cappelli. F; Arukwe. A. (2019). Biotransformation and oxidative stress responses in rat
hepatic cell-line (H4IIE) exposed to organophosphate esters (OPEs). Toxicol Appl Pharmacol
371: 84-94. http://dx.doi.Org/10.1016/i.taap.2019.04.004
Meyer. J: Bester. K. (2004). Organophosphate flame retardants and plasticisers in wastewater treatment
plants. J Environ Monit 6: 599-605. http://dx.doi.org/10.1039/b403206c
Meyer. RL: Balgooven. TG. (1987). A study and implications of habitat separation by sex of wintering
American kestrels (Falco sparverius L.). Raptor Research 6: 107-123.
Meylan. WM: Howard. PH. (1991). Bond contribution method for estimating Henry's law constants.
Environ Toxicol Chem 10: 1283-1293. http://dx.doi.org/10.1002/etc.56201010Q7
Page 417 of 638
-------
Mihailovic. I; Fries. E. (2012). Atmospheric deposition of chlorinated organophosphate flame retardants
(OFR) onto soils. Atmos Environ 56: 177-183. http://dx.doi.Org/10.1016/i.atmosenv.2012.03.054
Mihailovic. I; Miloradov. MY: Fries. E. (2011). Application of Twisselmann extraction, SPME, and
GC-MS to assess input sources for organophosphate esters into soil. Environ Sci Technol 45:
2264-2269. http://dx.doi.org/10.1021/esl03870f
Minegishi. KI; Kurebavashi. H: Nambaru. S: Morimoto. K; Takahashi. T; Yamaha. T. (1988).
Comparative studies on absorption, distribution, and excretion of flame retardants halogenated
alkyl phosphate in rats. Eisei Kagaku 34: 102-114. http://dx.doi.org/10.1248/ihsl956.34.102
Mokra. K; Bukowski. K; Wozniak. K. (2018). Effects of tris(l-chloro-2-propyl)phosphate and tris(2-
chloroethyl)phosphate on cell viability and morphological changes in peripheral blood
mononuclear cells (in vitro study). Hum Exp Toxicol 37: 1336-1345.
http://dx.doi.org/10.1177/0960327118783529
Moller. A: Sturm. R; Xie. Z; Cai. M; He. J: Ebinghaus. R. (2012). Organophosphorus flame retardants
and plasticizers in airborne particles over the Northern Pacific and Indian Ocean toward the Polar
Regions: Evidence for global occurrence. Environ Sci Technol 46: 3127-3134.
http ://dx.doi.org/10.1021/es204272v
Moller. A: Xie. Z; Caba. A: Sturm. R; Ebinghaus. R. (2011). Organophosphorus flame retardants and
plasticizers in the atmosphere of the North Sea. Environ Pollut 159: 3660-3665.
http: //dx. doi. or g/10.1016/i. envpol .2011.07.022
Moran. IL; Tidwell. L; Barton. M; Kile. M; Miller. P; Rohlman. D; Seguinot-Medina. S: Ungwiluk. B;
Waghiyi. V: Anderson. K. (2023). Diffusive fluxes of persistent organic pollutants between
Arctic atmosphere, surface waters and sediments. Sci Total Environ 892: 164566.
http: //dx. doi. or g/10.1016/i. scitotenv .2023.164566
Moser. VC: Phillips. PM; Hedge. JM; McDaniel. KL. (2015). Neurotoxicological and thyroid
evaluations of rats developmentally exposed to tris(l,3-dichloro-2-propyl)phosphate (TDCIPP)
and tris(2-chloro-2-ethyl)phosphate (TCEP). Neurotoxicol Teratol 52: 236-247.
http://dx.doi.Org/10.1016/i.ntt.2015.08.004
Muir. DCG. (1984). Phosphate esters. In J Schoer (Ed.), (pp. 41-66). Berlin, Germany: Springer-Verlag.
http://dx.doi.org/10.1007/978-3-540-38819-7 2
Na. G: Hou. C: Li. R; Shi. Y; Gao. H; Jin. S: Gao. Y; Jiao. L; Cai. Y. (2020). Occurrence, distribution,
air-seawater exchange and atmospheric deposition of organophosphate esters (OPEs) from the
Northwestern Pacific to the Arctic Ocean. Mar Pollut Bull 157: 111243.
http://dx.doi.Org/10.1016/i.marpolbul.2020.l 11243
Nakamura. A: Tateno. N: Koiima. S: Kaniwa. MA: Kawamura. T. (1979). The mutagenicity of
halogenated alkanols and their phosphoric acid esters for Salmonella typhimurium. Mutat Res
66: 373-380. http://dx.doi.org/10.1016/0165-1218(79)90048-X
NASA. (2020). Comment submitted by Denise Thaller, Director, Environmental Management Division,
National Aeronautics and Space Administration (NASA) regarding draft scopes of the risk
evaluations to be conducted for seven chemical substances under the Toxic Substance Control
Act (TSCA). (EPA-HQ-OPPT-2018-0438-0057). Washington, DC.
https://www.regulations.gov/document/EPA-HO-OPPT-2018-0438-0Q57
Navaranian. G: Jantunen. LM; Diamond. ML: Harris. SA; Bernstein. S: Scott. JA; Takaro. T; Dai. R;
Lefebvre. PL: Mandhane. PJ: Moraes. TJ; Simons. E; Turvev. SE; Sears. MR: Subbarao. P;
Brook. J. R. (2021). Early life exposure to tris(2-butoxyethyl) phosphate (TBOEP) is related to
the development of childhood asthma. Environ Sci Technol Lett 8: 531-537.
http ://dx.doi.org/10.1021/acs.estlett.lc00210
NCBI. (2020). PubChem database: Compound summary: Tris(2-chloroethyl) phosphate.
https://pubchem.ncbi.nlm.nih.gov/compound/Tris 2-chloroethyl -phosphate
NFPA. (2022). US Fire Department Profile 2020. Washington, DC.
Page 418 of 638
-------
NICNAS. (2001). Priority existing chemical assessment report no. 17: Trisphosphates. (PEC 17).
Sydney, Australia. https://www.industrialchemicals.gov.au/sites/default/files/PEC17-
Trisphosphates.pdf
NIOSH. (2018). Health Hazard Evaluation Report: HHE-2015-0050-3308, May 2018. Evaluation of
Exposure to Metals, Flame Retardants, and Nanomaterials at an Electronics Recycling Company.
Beaucham, CC; Ceballos, D; Page, EH; Mueller, C; Calafat, AM; Sjodin, A; Ospina, M; La
Guardia, M; Glassford, E.
NIVA. (2008). Screening of selected metals and new organic contaminants 2007: Phosphorus flame
retardents, polyfluorinated organic compounds, nitro-PAHs, silver, platinum and sucralose in air,
wastewater treatment falcilities, and freshwater and marine recipients. (1014/2008; TA-
2367/2008). Oslo, Norway, https://niva.brage.unit.no/niva-xmlui/handle/11250/213998
NLM. (2015). PubChem: Hazardous substance data bank: Tris(2-chloroethyl) phosphate, 115-96-8.
Available online at
NLM. (2019). PubChem Database: Tris(2-chloroethyl) phosphate, CID=8295. National Library of
Medicine, National Center for Biotechnology Information.
https://pubchem.ncbi.nlm.nih.gov/compound/8295
Noack. (1993). Unveroffentl. Unters. im auftrag der Hoechts AG [Unpublished investigation on behalf
of Hoechts AG], (Test No. AAE37081, Project No. 930818HH). Noack.
NTP. (1991a). Final report on the reproductive toxicity of tris(2-chloroethyl) phosphate in CD-I Swiss
mice. (RACB92040). Research Triangle Park, NC: National Institute of Environmental Health
Sciences, National Toxicology Program.
https://ntrl.ntis.gov/NTRL/dashboard/searchResults/titleDetail/PB92129170.xhtml
NTP. (1991b). Toxicology and carcinogenesis studies of tris(2-chloroethyl) phosphate (CAS No. 115-
96-8) in F344/N rats and B6C3F1 mice (gavage studies) [NTP] (pp. 1-233). (TR-391). Research
Triangle Park, NC: U.S. Department of Health and Human Services, National Institutes of
Health, National Toxicology Program.
https://search.proquest.eom/docview/l 859394969?accountid=l 71501
NTP. (1992). CAMEO Chemicals Database of Hazardous Materials [Website],
http://cameochemicals.noaa. gov./
ODPHP. (2023a). Healthy People 2030 - Social determinants of health literature summaries:
Neighborhood and built environment [Website],
ODPHP. (2023b). Healthy People 2030 - Social determinants of health literature summaries: Poverty
[Website], https://health.gov/healthvpeople/prioritv-areas/social-determinants-health/literature-
summaries/poverty
ODPHP. (2023c). Healthy People 2030 - Social determinants of health literature summaries: Social and
community context [Website],
OECD. (2006). SIDS Initial Assessment Profile for SIAM 23: Tris(2-chloroethyl)phosphate (CAS no.
115-96-8) [OECD SIDS], Paris, France: Organization for Economic Co-operation and
Development. https://hpvchemicals.oecd.org/UI/handler.axd?id=bl86df35-d9cc-4767-b218-
603aelc3bll9
OECD. (2007). Detailed review paper on cell transformation assays for detection of chemical
carcinogens. In Series on Testing and Assessment No 31. (ENV/JM/MONO(2007)18). Paris,
France.
http://www.oecd. org/officialdocuments/publicdisplavdocumentpdf/?doclanguage=en&cote=env/
im/mono(2007)18
OECD. (2009). Emission scenario document on plastic additives. (JT03267870). Paris, France: OECD
Environmental Health and Safety Publications.
http://www.oecd. org/officialdocuments/publicdisplavdocumentpdf/?cote=env/im/mono(2004)8/r
evl&doclanguage=en
Page 419 of 638
-------
OECD. (2012). Test No. 305: Bioaccumulation in fish: Aqueous and dietary exposure. In
9789264185296. Paris, France: OECD Publishing, http://dx.doi.org/10.1787/9789264185296-en
OECD. (2016a). Test No. 473: In vitro mammalian chromosomal aberration test. Paris, France.
http://dx.doi.org/10.1787/9789264264649-en
OECD. (2016b). Test no. 474: Mammalian erythrocyte micronucleus test [Standard], In OECD
Guidelines for the Testing of Chemicals. Paris, France: OECD Publishing.
http://dx.doi.org/10.1787/9789264264762-en
OECD. (2016c). Test No. 476: In Vitro Mammalian Cell Gene Mutation Tests using the Hprt and xprt
genes. Paris, France, http://dx.doi.org/10.1787/9789264264809-en
OECD. (2017). Overview of the set of OECD Genetic Toxicology Test Guidelines and updates
performed in 2014-2015. Series on Testing & Assessment No. 238 - 2nd edition.
(ENV/JM/MONO(2016)33/REV1). Paris, France.
http://www.oecd.org/officialdocuments/publicdisplavdocumentpdf
OECD. (2022). Series on Testing & Assessment, No. 156: Guidance notes on dermal absorption studies
(Second edition). (ENV/JM/MONO(2011)36/REV1). Paris, France: Organisation for Economic
Co-operation and Development (OECD).
https://www.oecd. org/officialdocuments/publicdisplavdocumentpdf/?cote=ENV-JM-
M0N0(2Q 11)3 6%20&doclanguage=en
Oh. J: Buckley. JP; Li. X: Gachigi. KK; Kannan. K; Lyu. WJ: Ames. JL; Barrett. ES: Bastain. TM;
Breton. CV: Buss. C: Croen. LA: Dunlop. AL; Ferrara. A: Ghassabian. A: Herbstman. JB;
Hernandez-Castro. I; Hertz-Picciotto. I: Kahn. LG: Karagas. MR: Kuiper. JR; McEvov. CT;
Meeker. JD; Morello-Frosch. R; Padula. AM: Romano. ME: Sathyanaravana. S: Schantz. S:
Schmidt. RJ; Simhan. H; Starling. AP; Tylavsky. FA: Yolk. HE: Woodruff. TJ; Zhu. YY;
Bennett. DH. (2024). Associations of organophosphate ester flame retardant exposures during
pregnancy with gestational duration and fetal growth: The Environmental Influences on Child
Health Outcomes (ECHO) program. Environ Health Perspect 132: 017004.
http://dx.doi.org/10.1289/ehpl3182
Okeme. JO. (2018). Examining the gas-particle partitioning of organophosphate esters: How reliable are
air measurements? Environ Sci Technol 52: 13834-13844.
http://dx.doi.org/10.1021/acs.est.8b04588
Okeme. JO: Nguyen. LV; Lorenzo. M; Dhal. S: Pico. Y; Arrandale. VH; Diamond. ML. (2018).
Polydimethylsiloxane (silicone rubber) brooch as a personal passive air sampler for semi-volatile
organic compounds. Chemosphere 208: 1002-1007.
http://dx.doi.Org/10.1016/i.chemosphere.2018.05.196
Okeme. JO: Rodgers. TFM; Parnis. JM; Diamond. ML: Bidleman. TF; Jantunen. LM. (2020). Gas
chromatographic estimation of vapor pressures and octanol-air partition coefficients of
semivolatile organic compounds of emerging concern. Journal of Chemical and Engineering
Data 65: 2467-2475. http://dx.doi.org/10.1021/acs.iced.9b01126
Ortiz Carrizales. YP. (2018) Factors influencing human exposure assessment of organophosphorus
flame retardants (OPFRS) via indoor dust ingestion. (Doctoral Dissertation). University of
Birmingham, Birmingham, UK. Retrieved from https://etheses.bham.ac.uk/id/eprint/8071/
Paciorek. KJ: Kratzer. RH: Kaufman. J: Nakahara. JH: Christos. T; Hartstein. AM. (1978). Thermal
oxidative degradation studies of phosphate esters. Am Ind Hyg Assoc J 39: 633.
http://dx.doi.org/10.1080/00028897785Q7827
Padhye. LP: Yao. H: Kung'u. FT: Huang. CH. (2014). Year-long evaluation on the occurrence and fate
of pharmaceuticals, personal care products, and endocrine disrupting chemicals in an urban
drinking water treatment plant. Water Res 51: 266-276.
http://dx.doi.Org/10.1016/i.watres.2013.10.070
Page 420 of 638
-------
Pang. L; Ge. L; Yang. P; He. H; Zhang. H. (2018). Degradation of organophosphate esters in sewage
sludge: Effects of aerobic/anaerobic treatments and bacterial community compositions.
Bioresour Technol 255: 16-21. http://dx.doi.Org/10.1016/i.biortech.2018.01.104
Pang. L; Yang. PJ; Zhao. JH; Zhang. HZ. (2016). Comparison of wastewater treatment processes on the
removal efficiency of organophosphate esters. Water Sci Technol 74: 1602-1609.
http://dx.doi.org/10.2166/wst.2016.356
Park. H; Choo. G: Kim. H; Oh. JE. (2018). Evaluation of the current contamination status of PFASs and
OPFRs in South Korean tap water associated with its origin. Sci Total Environ 634: 1505-1512.
http: //dx. doi. or g/10.1016/i. scitotenv .2018.04.068
Percy. Z; Chen. A: Yang. W: Braun. JM; Lanphear. B; Ospina. M; Calafat. AM: Xie. C: Cecil. KM:
Vuong. AM: Xu. Y; Yolton. K. (2022). Childhood urinary organophosphate esters and cognitive
abilities in a longitudinal cohort study. Environ Res 215:114265.
http: //dx. doi. or g/10.1016/i. envres .2022.114265
Percy. Z; Vuong. AM: Xu. Y; Xie. C: Ospina. M; Calafat. AM: Lanphear. BP: Braun. JM: Cecil. KM:
Dietrich. KN: Chen. A: Yolton. K. (2021). Prenatal exposure to a mixture of organophosphate
esters and intelligence among 8-year-old children of the HOME Study. Neurotoxicology 87:
149-155. http://dx.doi.Org/10.1016/i.neuro.2021.09.005
Peverly. AA: Ma. Y; Venier. M: Rodenburg. Z: Spak. SN: Hornbuckle. KC: Hites. RA. (2015).
Variations of flame retardant, polycyclic aromatic hydrocarbon, and pesticide concentrations in
Chicago's atmosphere measured using passive sampling. Environ Sci Technol 49: 5371-5379.
http://dx.doi.org/10.1021/acs.est.5b00216
Phillips. AL: Hammel. SC: Hoffman. K: Lorenzo. AM: Chen. A: Webster. TF: Stapleton. HM. (2018).
Children's residential exposure to organophosphate ester flame retardants and plasticizers:
Investigating exposure pathways in the TESIE study. Environ Int 116: 176-185.
http://dx.doi.Org/10.1016/i.envint.2018.04.013
Poet. TS: Corlev. RA: Thrall. KD: Edwards. JA: Tanoio. H: Weitz. KK: Hui. X: Maibach. HI: Wester.
RC. (2000). Assessment of the percutaneous absorption of trichloroethylene in rats and humans
using MS/MS real-time breath analysis and physiologically based pharmacokinetic modeling.
Toxicol Sci 56: 61-72. http://dx.doi.Org/10.1093/toxsci/56.l.61
Poma. G: Sales. C: Bruvland. B: Christia. C: Goscinny. S: Van Loco. J: Covaci. A. (2018). Occurrence
of organophosphorus flame retardants and plasticizers (PFRs) in Belgian foodstuffs and
estimation of the dietary exposure of the adult population. Environ Sci Technol 52: 2331-2338.
http://dx.doi.org/10.1021/acs.est.7b06395
PPG. (2010). Safety data sheet: PITT-CHAR XP EP 97-194 component A. Borgerhout, Belgium: PPG
Coatings SPRL/BVBA.
PPG. (2016). Safety data sheet: PITT-CHAR XP part A off white. Pittsburgh, PA: PPG Industries, Inc.
Price. K: Haddad. S: Krishnan. K. (2003). Physiological modeling of age-specific changes in the
pharmacokinetics of organic chemicals in children. J Toxicol Environ Health A 66: 417-433.
http://dx.doi.org/10.1080/1528739030645Q
Prival. MJ: McCoy. EC: Gutter. B: Rosenkranz. HS. (1977). Tris(2,3-dibromopropyl) phosphate:
Mutagenicity of a widely used flame retardant. Science 195: 76-78.
http://dx.doi.org/10.1126/science.318761
Propp. VR: De Silva. AO: Spencer. C: Brown. SJ: Catingan. SD: Smith. JE: Roy. JW. (2021). Organic
contaminants of emerging concern in leachate of historic municipal landfills. Environ Pollut 276:
116474. http://dx.doi.Org/10.1016/i.envpol.2021.l 16474
Oi. C: Yu. G: Zhong. M: Peng. G: Huang. J: Wang. B. (2019). Organophosphate flame retardants in
leachates from six municipal landfills across China. Chemosphere 218: 836-844.
http: //dx. doi. or g/10.1016/i. chemosphere .2018.11.150
Page 421 of 638
-------
Raimondo. S; Lilavois. C; Nelson. SA. (2023). Uncertainty analysis and updated user guidance for
interspecies correlation estimation models and low toxicity compounds. Integr Environ Assess
Manag. http://dx.doi.org/10.1002/ieam.4884
R.ai m on do. SDNV; Barron. MG. (2010). Web-Based Interspecies Correlation Estimation (Web-ICE) for
Acute Toxicity: User Manual Version 3.1. (600R10004). Raimondo, S., D.N. Vivian, and M.G.
Barron, http://nepis.epa.gov/exe/ZvPURL.cgi?Dockev=P 10068ND.txt
Rantakokko. P; Kumar. E; Braber. J: Huang. T; Kiviranta. H; Cequier. E; Thomsen. C. (2019).
Concentrations of brominated and phosphorous flame retardants in Finnish house dust and
insights into children's exposure. Chemosphere 223: 99-107.
http: //dx. doi. or g/10.1016/i. chemosphere .2019.02.027
Rauert. C: Schuster. JK; Eng. A: Harner. T. (2018). Global atmospheric concentrations of brominated
and chlorinated flame retardants and organophosphate esters. Environ Sci Technol 52: 2777-
2789. http://dx.doi.org/10.1021/acs.est.7b06239
Reemtsma. T; Berger. U: Arp. HPH; Gallard. H; Knepper. TP: Neumann. M; Quintana. JB; de Voogt. P.
(2016). Mind the gap: Persistent and mobile organic compounds—Water contaminants that slip
through. Environ Sci Technol 50: 10308-10315. http://dx.doi.org/10.1021/acs.est.6b03338
Reers. AR; Eng. ML: Williams. TP: Elliott. JE; Cox. ME: Beischlag. TV. (2016). The Flame-Retardant
Tris( 1,3-dichloro-2-propyl) Phosphate Represses Androgen Signaling in Human Prostate Cancer
Cell Lines. J Biochem Mol Toxicol 30: 249-257. http://dx.doi.org/10.1002/ibt.21786
Regnery. J: Puettmann. W. (2009). Organophosphorus flame retardants and plasticizers in rain and snow
from middle Germany. CLEAN - Soil, Air, Water 37: 334-342.
http://dx.doi.org/10.1002/clen.20090005Q
Regnery. J: Piittmann. W. (2010a). Occurrence and fate of organophosphorus flame retardants and
plasticizers in urban and remote surface waters in Germany. Water Res 44: 4097-4104.
http://dx.doi.Org/10.1016/i.watres.2010.05.024
Regnery. J: Piittmann. W. (2010b). Seasonal fluctuations of organophosphate concentrations in
precipitation and storm water runoff. Chemosphere 78: 958-964.
http://dx.doi.Org/10.1016/i.chemosphere.2009.12.027
Ren. G: Chu. X: Zhang. J: Zheng. K; Zhou. X: Zeng. X: Yu. Z. (2019). Organophosphate esters in the
water, sediments, surface soils, and tree bark surrounding a manufacturing plant in north China.
Environ Pollut 246: 374-380. http://dx.doi.Org/10.1016/i.envpol.2018.12.020
Ren. X: Han. HJ; Lee. YJ; Lee. SH; Ng. HY; Chae. KJ; Kim. IS. (2012). Proapoptotic effect of a
micropollutant (tris-(2-chloroethyl)-phosphate) at environmental level in primary cultured renal
proximal tubule cells. J Water Health 10: 522-530. http://dx.doi.org/10.2166/wh.2012.180
Ren. X: Lee. YJ: Han. HJ: Kim. IS. (2008). Effect of tris-(2-chloroethyl)-phosphate (TCEP) at
environmental concentration on the levels of cell cycle regulatory protein expression in primary
cultured rabbit renal proximal tubule cells. Chemosphere 74: 84-88.
http://dx.doi.Org/10.1016/i.chemosphere.2008.09.022
Ren. X: Lee. YJ: Han. HJ: Kim. IS. (2009). Cellular effect evaluation of micropollutants using
transporter functions of renal proximal tubule cells. Chemosphere 77: 968-974.
http://dx.doi.Org/10.1016/i.chemosphere.2009.08.012
Reynolds. CW: Foon. KA. (1984). T gamma-lymphoproliferative disease and related disorders in
humans and experimental animals: A review of the clinical, cellular, and functional
characteristics [Review], Blood 64: 1146-1158.
http://dx.doi.Org/10.l 182/blood.V64.6.1146.1146
Ridolfi. (2016). Heritage fish consumption rates of the Kootenai Tribe of Idaho. Washington, DC: U.S.
Environmental Protection Agency, https://www.epa.gov/sites/default/files/2017-
01/ documents/heritage-fi sh-consumption-rates-kootenai-dec2016. pdf
Page 422 of 638
-------
Rodgers. TFM; Truong. JW; Jantunen. LM; Helm. PA; Diamond. ML. (2018). Organophosphate Ester
Transport, Fate, and Emissions in Toronto, Canada, Estimated Using an Updated Multimedia
Urban Model. Environ Sci Technol 52: 12465-12474. http://dx.doi.org/10.1021/acs.est.8b02576
SAB. (2007). Science Advisory Board (SAB) review of the Estimation Programs Interface Suite (EPI
Suite). (EPA-SAB-07-11). Washington, DC: U.S. Environmental Protection Agency.
Saini. A: Thavsen. C: Jantunen. L; McQueen. RH; Diamond. ML. (2016). From clothing to laundry
water: Investigating the fate of phthalates, brominated flame retardants, and organophosphate
esters. Environ Sci Technol 50: 9289-9297. http://dx.doi.org/10.1021/acs.est.6b02038
Saint-Hilaire. D; Ismail. KZ; Jans. U. (2011). Reaction of tris(2-chloroethyl)phosphate with reduced
sulfur species. Chemosphere 83: 941-947. http://dx.doi.org/10.1016/i.chemosphere.2011.02.040
Sala. B; Gimenez. J: de Stephanis. R; Barcelo. D; Eliarrat. E. (2019). First determination of high levels
of organophosphorus flame retardants and plasticizers in dolphins from Southern European
waters. Environ Res 172: 289-295. http://dx.doi.Org/10.1016/i.envres.2019.02.027
Sala. M; Gu. ZG: Moens. G: Chouroulinkov. I. (1982). In vivo and in vitro biological effects of the
flame retardants tris(2,3-dibromopropyl) phosphate and tris(2-chlorethyl)orthophosphate. Eur J
Cancer Clin Oncol 18: 1337-1344. http://dx.doi.org/10.1016/0277-5379(82)90138-9
Salamova. A: Ma. Y; Venier. M; Hites. RA. (2014). High levels of organophosphate flame retardants in
the great lakes atmosphere. Environ Sci Technol Lett 1: 8-14.
http://dx.doi.org/10.1021/ez400Q34n
Salamova. A: Peverly. AA; Venier. M; Hites. RA. (2016). Spatial and temporal trends of particle phase
organophosphate ester concentrations in the atmosphere of the great lakes. Environ Sci Technol
50: 13249-13255. http://dx.doi.org/10.1021/acs.est.6b04789
Sasaki. K; Suzuki. T; Takeda. M; Uchiyama. M. (1982). Bioconcentration and excretion of phosphoric
acid triesters by killifish (Oryzeas latipes). Bull Environ Contam Toxicol 28: 752-759.
http://dx.doi.org/10.1007/BF016Q5648
Scheller. J: Chalaris. A: Schmidt-Arras. D; Rose-John. S. (2011). The pro- and anti-inflammatory
properties of the cytokine interleukin-6 [Review], Biochim Biophys Acta 1813: 878-888.
http://dx.doi.Org/10.1016/i.bbamcr.2011.01.034
Schenk. A: Ghallab. A: Hofmann. U; Hassan. R; Schwarz. M; Schuppert. A: Schwen. LO; Braeuning.
A: Teutonico. D; Hengstler. JG: Kuepfer. L. (2017). Physiologically-based modelling in mice
suggests an aggravated loss of clearance capacity after toxic liver damage. Sci Rep 7: 6224.
http://dx.doi. org/10.103 8/s41598-017-045 74-z
SCHER. (2012). Opinion on tris(2-chloroethyl)phosphate (TCEP) in toys. European Commission
Scientific Committee on Health and Environmental Risks.
http://ec.europa.eu/health/scientific committees/environmental risks/docs/scher o 158.pdf
Schreder. ED: La Guardia. MJ. (2014). Flame retardant transfers from U.S. households (dust and
laundry wastewater) to the aquatic environment. Environ Sci Technol 48: 11575-11583.
http://dx.doi.org/10.1021/es502227h
Schreder. ED: Uding. N: La Guardia. MJ. (2016). Inhalation a significant exposure route for chlorinated
organophosphate flame retardants. Chemosphere 150: 499-504.
http://dx.doi.Org/10.1016/i.chemosphere.2015.l 1.084
Scott. BF; Sverko. E; Maguire. RJ. (1996). Determination of benzothiazole and alkylphosphates in water
samples from the Great Lakes drainage basin by gas chromatography/atomic emission detection.
Water Qual Res J Can 31: 341-360. http://dx.doi.org/10.2166/wqri.1996.021
Sengupta. A: Lyons. JM; Smith. DJ; Drewes. JE; Snyder. SA; Heil. A: Maruva. KA. (2014). The
occurrence and fate of chemicals of emerging concern in coastal urban rivers receiving discharge
of treated municipal wastewater effluent. Environ Toxicol Chem 33: 350-358.
http: //dx. doi. or g/10.1002/etc.245 7
Page 423 of 638
-------
Sharma. R; Sharma. S. (2023). StatPearls: Physiology, blood volume. Treasure Island, FL: StatPearls
Publishing. https://www.ncbi.nlm.nih.gov/books/NBK526077/
Sharma. U; Pal D; Prasad. R. (2014). Alkaline phosphatase: An overview [Review], Indian J Clin
Biochem 29: 269-278. http://dx.doi.org/10.1007/sl2291-013-04Q8-v
Shen. B; Whitehead. TP: Gill R; Dhaliwal. J: Brown. FR; Petreas. M; Patton. S: Hammond. SK. (2018).
Organophosphate flame retardants in dust collected from United States fire stations. Environ Int
112: 41-48. http://dx.doi.Org/10.1016/i.envint.2017.12.009
Shepel'skaia. NR; Dyshginevich. NE. (1981). [Experimental study of the gonadotoxic effect of tri-
(chloroethyl)-phosphate]. Gig Sanit20-21.
Shin. HM; McKone. TE; Nishioka. MG: Fallin. MP: Croen. LA: Hertz-Picciotto. I: Newschaffer. CJ:
Bennett. DH. (2014). Determining source strength of semivolatile organic compounds using
measured concentrations in indoor dust. Indoor Air 24: 260-271.
http://dx.doi.org/10. Ill 1/ina. 12070
Shoeib. M; Ahrens. L; Jantunen. L; Harner. T. (2014). Concentrations in air of organobromine,
organochlorine and organophosphate flame retardants in Toronto, Canada. Atmos Environ 99:
140-147. http://dx.doi.Org/10.1016/i.atmosenv.2014.09.040
Sigma-Aldrich. (2019). Tris(2-chloroethyl) phosphate. Sigma-Aldrich Inc.
https://www.sigmaaldrich.eom/catalog/product/aldrich/l 19660?lang=en®ion=US
Simmon. VF; Kauhanen. K; Tardiff. RG. (1977). Mutagenic activity of chemicals identified in drinking
water. In D Scott; B Bridges; F Sobel (Eds.), Developments in Toxicology and Environmental
Science, 2 (pp. 249-258). New York, NY: Elsevier/North Holland Press.
Siodin. A; Carlsson. H; Thuresson. K; Siolin. S; Bergman. A; Ostman. C. (2001). Flame retardants in
indoor air at an electronics recycling plant and at other work environments. Environ Sci Technol
35: 448-454. http://dx.doi.org/10.1021/esQ00077n
Snyder. SA; Wert. EC; Rexing. DJ; Zegers. RE; Drury. DP. (2006). Ozone Oxidation of Endocrine
Disruptors and Pharmaceuticals in Surface Water and Wastewater. Ozone: Science and
Engineering 28: 445-460. http://dx.doi.org/10.1080/01919510601Q39726
Sonich-Mullin. C; Fielder. R; Wiltse. J; Baetcke. K; Dempsev. J; Fenner-Crisp. P; Grant. D; Hartley. M;
Knaap. A; Kroese. D; Mangelsdorf. I; Meek. E; Rice. JM; Younes. M. (2001). IPCS conceptual
framework for evaluating a mode of action for chemical carcinogenesis. Regul Toxicol
Pharmacol 34: 146-152. http://dx.doi.org/10.1006/rtph.2001.1493
Stachel. B; Jantzen. E; Knoth. W; Kruger. F; Lepom. P; Oetken. M; Reincke. H; Sawal. G; Schwartz. R;
Uhlig. S. (2005). The Elbe Flood in August 2002—Organic Contaminants in Sediment Samples
Taken After the Flood Event. J Environ Sci Health A Tox Hazard Subst Environ Eng 40: 265-
287. http://dx.doi.org/10.1081/ESE-200Q45531
Stackelberg. PE; Furlong. ET; Meyer. MT; Zaugg. SD; Henderson. AK; and Reissman. DB. (2004).
Persistence of pharmaceuticals compounds and other organic wastewater contaminants in a
conventional drinking-water-treatment plant. Sci Total Environ 329: 99-113.
http: //dx. doi. or g/10.1016/i. scitotenv .2004.03.015
Stapleton. HM; Klosterhaus. S; Keller. A; Ferguson. P; van Bergen. S; Cooper. E; Webster. TF; Blum.
A. (2011). Identification of Flame Retardants in Polyurethane Foam Collected from Baby
Products. Environ Sci Technol 45: 5323-5331. http://dx.doi.org/10.1021/es2007462
Stapleton. HM; Sharma. S; Getzinger. G; Ferguson. P; Gabriel. M; Webster. TF; Blum. A. (2012).
Novel and High Volume Use Flame Retardants in US Couches Reflective of the 2005 PentaBDE
Phase Out. Environ Sci Technol 46: 13432-13439. http://dx.doi.org/10.1021/es303471d
Stubbings. WA; Guo. JH; Simon. K; Romanak. K; Bowerman. W; Venier. M. (2018). Flame retardant
metabolites in addled bald eagle eggs from the Great Lakes region. Environ Sci Technol Lett 5:
354-359. http://dx.doi.org/10.1021/acs.estlett.8b00163
Page 424 of 638
-------
Stubbings. WA; Nguyen. LV; Romanak. K; Jantunen. L; Melymuk. L; Arrandale. V; Diamond. ML;
Venier. M. (2019). Flame retardants and plasticizers in a Canadian waste electrical and electronic
equipment (WEEE) dismantling facility. Sci Total Environ 675: 594-603.
Su. G; Letcher. RJ; Yu. H. (2016). Organophosphate flame retardants and plasticizers in aqueous
solution: pH-dependent hydrolysis, kinetics, and pathways. Environ Sci Technol 50: 8103-8111.
http://dx.doi.org/10.1021/acs.est.6b02187
Stihring. R; Diamond. ML: Scheringer. M; Wong. F; Pucko. M; Stern. G: Burt. A: Hung. H; Fellin. P;
Li. Ft; Jantunen. LM. (2016). Organophosphate esters in Canadian Arctic air: Occurrence, levels
and trends. Environ Sci Technol 50: 7409. http://dx.doi.org/10.1021/acs.est.6b00365
Sun. L; Tan. H; Peng. T; Wang. S; Xu. W; Qian. H; Jin. Y; Fu. Z. (2016). Developmental neurotoxicity
of organophosphate flame retardants in early life stages of Japanese medaka (Oryzias latipes).
Environ Toxicol Chem 35: 2931-2940. http://dx.doi.org/10.1002/etc.3477
Sundkvist. AM; Olofsson. U; Haglund. P. (2010). Organophosphorus flame retardants and plasticizers in
marine and fresh water biota and in human milk. J Environ Monit 12: 943-951.
http://dx.doi.org/10.1039/b921910b
Sutha. J; Anila. PA; Umamaheswari. S; Ramesh. M; Naravanasamy. A; Poopal. RK; Ren. Z. (2020).
Biochemical responses of a freshwater fish Cirrhinus mrigala exposed to tris(2-chloroethyl)
phosphate (TCEP). Environ Sci Pollut Res Int 27: 34369-34387.
http://dx.doi. org/10.1007/s 113 56-020-09527-0
Sutton. R; Chen. D; Sun. J; Greig. DJ; Wu. Y. (2019). Characterization of brominated, chlorinated, and
phosphate flame retardants in San Francisco Bay, an urban estuary. Sci Total Environ 652: 212-
223. http://dx.doi.org/10.1016/i. scitotenv.2018.10.096
Takada. H; Karapanagioti. H. (2019). Hazardous chemicals associated with plastics in the marine
environment. In The Handbook of Environmental Chemistry (HEC), vol 78. Cham, Switzerland:
Springer, http://dx.doi.org/10.1007/978-3-319-95568-l
Takada. K; Yasuhara. K; Nakaii. Y; Yoshimoto. H; Momma. J; Kurokawa. Y; Aida. Y; Tobe. M.
(1989). [Carcinogenicity study of tris (2-chloroethyl) phosphate in ddY mice], J Toxicol Pathol
2: 213-222. http://dx.doi.Org/10.1293/tox.2.213
Tamokou. J; Kuete. V. (2014). Chapter 10: Mutagenicity and carcinogenicity of African medicinal
plants. In V Kuete (Ed.), (pp. 277-322). Amsterdam, Netherlands: Elsevier.
http://dx.doi.org/10.1016/B978-0-12-800018-2.0001Q-8
Tan. H; Yang. L; Yu. Y; Guan. O; Liu. X; Li. L; Chen. D. (2019). Co-existence of organophosphate di-
and tri-esters in house dust from South China and Midwestern United States: Implications for
human exposure. Environ Sci Technol 53: 4784-4793. http://dx.doi.org/10.1021 /acs.est.9b00229
Tang. B; Poma. G; Bastiaensen. M; Yin. SS; Luo. XJ; Mai. BX; Covaci. A. (2019). Bioconcentration
and biotransformation of organophosphorus flame retardants (PFRs) in common carp (Cyprinus
carpio). Environ Int 126: 512-522. http://dx.doi.Org/10.1016/i.envint.2019.02.063
Taniai. E; Havashi. H; Yafune. A; Watanabe. M; Akane. H; Suzuki. K; Mitsumori. K; Shibutani. M.
(2012a). Cellular distribution of cell cycle-related molecules in the renal tubules of rats treated
with renal carcinogens for 28 days: Relationship between cell cycle aberration and
carcinogenesis. Arch Toxicol 86: 1453-1464. http://dx.doi.org/10.1007/s00204-012-Q829-z
Taniai. E; Yafune. A; Havashi. H; Itahashi. M; Hara-Kudo. Y; Suzuki. K; Mitsumori. K; Shibutani. M.
(2012b). Aberrant activation of ubiquitin D at G2 phase and apoptosis by carcinogens that evoke
cell proliferation after 28-day administration in rats. J Toxicol Sci 37: 1093-1 111.
http://dx.doi.org/10.2131/its.37.1093
TCI America. (2018). Safety Data Sheet: Tris(2-chloroethyl) phosphate, P0268. Portland, OR.
TERA. (2013). Toxicity review of tris(2-chloroethyl) phosphate (TCEP). U.S. Consumer Product Safety
Commission, Directorate for Hazard Identification and Reduction.
Page 425 of 638
-------
https://web.archive.Org/web/20190320055405/https://www.cpsc. gov/s3fs-public/pdfs/TCEP-
contract-report-with-cover-letter.pdf
TERA. (2015). Environmental concentrations and consumer exposure data for tris(2-chloroethyl)
phosphate (TCEP). Cincinnati, OH: Toxicology Excellence for Risk Assessment Center at the
University of Cincinnati.
https://web.archive.org/web/20170207003923/https://www.cpsc. gov/s3fs-
public/pdfs/CPSCStaffStatementToxicologyExcellenceRiskAssessmentsReportExposureDataTC
EP.pdf
Thomas. J: Haseman. JK; Goodman. JI; Ward. JM; Loughran. TP. Jr.: Spencer. PJ. (2007). A review of
large granular lymphocytic leukemia in Fischer 344 rats as an initial step toward evaluating the
Implication of the endpoint to human cancer risk assessment [Review], Toxicol Sci 99: 3-19.
http://dx.doi. org/10.1093/toxsci/kfm098
Tilson. HA: Veronesi. B; McLamb. RL; Matthews. HB. (1990). Acute exposure to tris(2-
chloroethyl)phosphate produces hippocampal neuronal loss and impairs learning in rats. Toxicol
Appl Pharmacol 106: 254-269. http://dx.doi.org/10.1016/0041 -008X(90)90245-P
Torav Research Center. (1997a). Acute immobilization test on Daphnia magna exposed to tris(2-
chloroethyl) phosphate (English translation). (NMMP/E97/2030). Tokyo, Japan: Environment
Agency of Japan.
Torav Research Center. (1997b). Algal growth inhibition test of Selenastrum capricornutum exposed to
tris(2-chloroethyl) phosphate (English translation). (NMMP/E97/1030). Tokyo, Japan:
Environment Agency of Japan.
Torav Research Center. (1997c). Reproductive inhibition test on Daphnia magna exposed to tris(2-
chloroethyl) phosphate (English translation). (NMMP/E97/3030). Tokyo, Japan: Environment
Agency of Japan.
Toscano. WA; Coleman. KP. (2012). Patty's Toxicology: Esters of carbonic and orthocarbonic acid,
organic phosphorous, monocarboxylic halogenated acids, haloalcohols, and organic silicon (6
ed.). New York, NY: John Wiley & Sons.
U.S. EPA. (1988). National survey of solid waste (municipal) landfill facilities [EPA Report],
(EPA/530-SW88-034). Washington, DC: Office of Solid Waste and Emergency Response, U.S.
Environmental Protection Agency.
U.S. EPA. (1991). Guidelines for developmental toxicity risk assessment. Fed Reg 56: 63798-63826.
U.S. EPA. (1992). Guidelines for exposure assessment. Federal Register 57(104):22888-22938 [EPA
Report], (EPA/600/Z-92/001). Washington, DC.
http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=15263
U.S. EPA. (1993a). Reference Dose (RfD): description and use in health risk assessments background
document 1A, March 15, 1993. Washington, DC: U.S. Environmental Protection Agency,
Integrated Risk Information System, https://www.epa.gov/iris/reference-dose-rfd-description-
and-use-health-risk-assessments
U.S. EPA. (1993b). Wildlife exposure factors handbook [EPA Report], (EPA/600/R-93/187).
Washington, DC: Office of Research and Development.
http://cfpub.epa. gov/ncea/cfm/recordisplav.cfm?deid=2799
U.S. EPA. (1996). Guidelines for reproductive toxicity risk assessment [EPA Report], (EPA/630/R-
96/009). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
https://nepis. epa.gov/Exe/ZyPURL. cgi?Dockev=3 0004YQB.txt
U.S. EPA. (1998a). Assessment of thyroid follicular cell tumors [EPA Report] (pp. 1-51). (EPA/630/R-
97/002). Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
https://www.epa.gov/sites/production/files/2014-ll/documents/thyroid.pdf
Page 426 of 638
-------
U.S. EPA. (1998b). Guidelines for ecological risk assessment [EPA Report], (EPA/630/R-95/002F).
Washington, DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
https://www.epa.gov/risk/guidelines-ecological-risk-assessment
U.S. EPA. (2000a). Interim guidance for using ready and inherent biodegradability tests to derive input
data for multimedia models and wastewater treatment plants (WWT) models (9/1/2000).
Washington, DC. https://www.epa.gov/tsca-screening-tools/interim-guidance-using-readv-and-
inherent-biodegradabilitv-tests-derive-0
U.S. EPA. (2000b). Methodology for deriving ambient water quality criteria for the protection of human
health (2000). (EPA/822/B-00/004). Washington, DC: U.S. Environmental Protection Agency,
Office of Water, http://www.epa.gov/waterscience/criteria/humanhealth/method/complete.pdf
U.S. EPA. (2000c). Supplementary guidance for conducting health risk assessment of chemical mixtures
(pp. 1-209). (EPA/630/R-00/002). Washington, DC: U.S. Environmental Protection Agency,
Risk Assessment Forum, http://cfpub.epa.gov/ncea/cfm/recordisplav.cfm?deid=20533
U.S. EPA. (2000d). The use of data on cholinesterase inhibition for risk assessments of
organophosphorous and carbamate pesticides. Washington, DC: U.S. Environmental Protection
Agency, Office of Pesticide Programs, https://www.epa.gov/sites/default/files/2015-
07/documents/cholin.pdf
U.S. EPA. (2002a). Guidance on cumulative risk assessment of pesticide chemicals that have a common
mechanism of toxicity [EPA Report], Washington, D.C.
U.S. EPA. (2002b). A review of the reference dose and reference concentration processes.
(EPA630P02002F). Washington, DC. https://www.epa.gov/sites/production/files/2014-
12/documents/rfd-final.pdf
U.S. EPA. (2003a). Attachment 1-3 Guidance for Developing Ecological Soil Screening Levels (Eco-
SSLs): Evaluation of Dermal Contact and Inhalation Exposure Pathways for the Purpose of
Setting Eco-SSLs. (OSWER9285755E). Washington, DC: .S. Environmental Protection Agency,
Office of Solid Waste and Emergency Response.
https://www.epa.gov/sites/production/files/2015-09/documents/ecossl attachment l-3.pdf
U.S. EPA. (2003b). Attachment 1-4. Guidance for developing ecological soil screening levels (Eco-
SSLs): Review of background concentration for metals. (OSWER Directive 92857-55).
Washington, DC. https://www.epa.gov/sites/default/files/2015-
09/documents/ecossl attachment l-4.pdf
U.S. EPA. (2003c). Methodology for deriving ambient water quality criteria for the protection of human
health (2000), technical support document. Volume 2: Development of national bioaccumulation
factors [EPA Report], (EPA/822/R-03/030). Washington, DC.
http://www.epa.gov/waterscience/criteria/humanhealth/method/tsdvol2.pdf
U.S. EPA. (2005a). Guidance for developing ecological soil screening levels [EPA Report], (OSWER
Directive 92857-55). Washington, DC: U.S. Environmental Protection Agency, Office of Solid
Waste and Emergency Response, http://www.epa.gov/chemical-research/guidance-developing-
ecological-soil-screening-levels
U.S. EPA. (2005b). Guidelines for carcinogen risk assessment [EPA Report], (EPA630P03001F).
Washington, DC. https://www.epa.gov/sites/production/files/2013-
09/documents/cancer guidelines final 3-25-05.pdf
U.S. EPA. (2005c). Supplemental guidance for assessing susceptibility from early-life exposure to
carcinogens [EPA Report], (EPA/630/R-03/003F). Washington, DC: U.S. Environmental
Protection Agency, Risk Assessment Forum, https://www.epa.gov/risk/supplemental-guidance-
assessing-susceptibilitv-early-life-exposure-carcinogens
U.S. EPA. (2007a). Attachment 4-3 Guidance for Developing Ecological Soil Screening Levels (Eco-
SSLs) Eco-SSL Standard Operating Procedure (SOP) #4: Wildlife Toxicity Reference Value
Page 427 of 638
-------
Literature Review, Data Extraction and Coding. (OSWER9285755F).
http://nepis.epa. gov/exe/ZvPURL.cgi?Dockev=P 1OOCDHC .txt
U.S. EPA. (2009). Provisional Peer-Reviewed Toxicity Values for tris(2-chloroethyl)phosphate (TCEP)
(CASRN 115-96-8) [EPA Report], (EPA/690/R-09/069F). Cincinnati, OH.
http://hhpprtv.ornl.gov/issue papers/Tris2chloroethylphosphate.pdf
U.S. EPA. (2011a). Exposure factors handbook: 2011 edition [EPA Report], (EPA/600/R-090/052F).
Washington, DC: U.S. Environmental Protection Agency, Office of Research and Development,
National Center for Environmental Assessment.
https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockev=P 100F2QS.txt
U.S. EPA. (201 lb). Recommended Use of Body Weight 3/4 as the Default Method in Derivation of the
Oral Reference Dose.
U.S. EPA. (2012a). Advances in inhalation gas dosimetry for derivation of a reference concentration
(RfC) and use in risk assessment (pp. 1-140). (EPA/600/R-12/044). Washington, DC.
https://cfpub.epa.gov/ncea/risk/recordisplav.cfm?deid=244650&CFID=50524762&CFTOKEN=
17139189
U.S. EPA. (2012b). Benchmark dose technical guidance [EPA Report], (EPA100R12001). Washington,
DC: U.S. Environmental Protection Agency, Risk Assessment Forum.
https://www.epa.gov/risk/benchmark-dose-technical-guidance
U.S. EPA. (2014a). Estimated fish consumption rates for the U.S. population and selected
subpopulations (NHANES 2003-2010) [EPA Report], (EPA-820-R-14-002). Washington, DC.
https://www.epa.gov/sites/production/files/2015-01/documents/fish-consumption-rates-2014.pdf
U.S. EPA. (2014b). Framework for human health risk assessment to inform decision making. Final
[EPA Report], (EPA/100/R-14/001). Washington, DC: U.S. Environmental Protection, Risk
Assessment Forum, https://www.epa.gov/risk/framework-human-health-risk-assessment-inform-
deci si on-making
U.S. EPA. (2015a). Flame retardants used in flexible polyurethane foam: An alternatives assessment
update. (744R15002). Washington, DC: U.S. Environmental Protection Agency, Design for the
Environment Program, https://www.epa.gov/sites/production/files/2015-
08/documents/ffr final.pdf
U.S. EPA. (2015b). TSCA Work Plan Chemical Problem Formulation and Initial Assessment.
Chlorinated Phosphate Ester Cluster Flame Retardants. (EPA 740-R1-5001). Washington, DC.
http://www.epa.gov/sites/production/files/2015-
09/documents/cpe fir cluster problem formulation.pdf
U.S. EPA. (2016a). Ecological effects test guidelines: OCSPP 850.1075: Freshwater and saltwater fish
acute toxicity test [EPA Report], (EPA 712-C-16-007; EPA-HQ-OPPT-2009-0154-0035).
Washington, DC: U.S. Environmental Protection Agency, Office of Chemical Safety and
Pollution Prevention. https://www.regulations.gov/document/EPA-HQ-OPPT-2009-0154-0035
U.S. EPA. (2016b). Guidance for conducting fish consumption surveys. (823B16002).
https://www.epa.gov/sites/production/files/2017-01/documents/fc survey guidance.pdf
U.S. EPA. (2016c). Integrated review plan for the national ambient air quality standards for particulate
matter [EPA Report], (EPA-452/R-16-005). Research Triangle Park, NC.
https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockev=P 100R5VE.txt
U.S. EPA. (2016d). Public database 2016 chemical data reporting (May 2017 release). Washington, DC:
US Environmental Protection Agency, Office of Pollution Prevention and Toxics.
https://www.epa.gov/chemical-data-reporting
U.S. EPA. (2016e). Weight of evidence in ecological assessment [EPA Report], (EPA/100/R-16/001).
Washington, DC: Office of the Science Advisor.
https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockev=P 100SFXR.txt
Page 428 of 638
-------
U.S. EPA. (2017a). Estimation Programs Interface Suite™ v.4.11. Washington, DC: U.S.
Environmental Protection Agency, Office of Pollution Prevention Toxics. Retrieved from
https://www.epa.gov/tsca-screening-tools/download-epi-suitetm-estimation-program-interface-
v411
U.S. EPA. (2017b). Human Health Benchmarks for Pesticides: Updated 2017 Technical Document (pp.
5). (EPA 822-R -17 -001). Washington, DC: U.S. Environmental Protection Agency, Office of
Water, https://www.epa.gov/sites/production/files/2015-10/documents/hh-benchmarks-
techdoc.pdf
U.S. EPA. (2017d). Update for Chapter 5 of the Exposure Factors Handbook: Soil and dust ingestion
[EPA Report], (EPA/600R-17/384F). Washington, DC: National Center for Environmental
Assessment, Office of Research and Development.
https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockev=P 100TTX4.txt
U.S. EPA. (2018). User's Guide for the AMS/EPA Regulatory Model (AERMOD). (EPA Document
Number: EPA-454/B-18-001). U.S. EPA.
U.S. EPA. (2019b). Chemistry dashboard information for tris(2-chloroethyl) phosphate. 115-96-8
[Website], https://comptox.epa.gov/dashboard/dsstoxdb/results?search=DTXSID5021411
U.S. EPA. (2019d). Consumer Exposure Model (CEM) 2.1 User Guide - Appendices. (EPA Contract #
EP-W-12-010). Washington, DC: U.S. EPA. https://www.epa.gov/tsca-screening-
tools/consumer-exposure-model-cem-version-20-users-guide
U.S. EPA. (2019e). Point Source Calculator: A Model for Estimating Chemical Concentration in Water
Bodies. Washington, DC: U.S. Environmental Protection Agency, Office of Chemical Safety and
Pollution Prevention.
U.S. EPA. (2019f). User's Guide: Integrated Indoor-Outdoor Air Calculator (IIOAC). Washington, DC:
U.S. EPA.
U.S. EPA. (2020b). Final Scope of the Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP);
CASRN 115-96-8. (EPA 740-R-20-009). Washington, DC: Office of Chemical Safety and
Pollution Prevention, https://www.epa.gov/sites/default/files/2020-09/documents/casrn 115-96-
8 tris2-chloroethyl phosphate tcep final scope.pdf
U.S. EPA. (2020c). Final Scope of the Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP);
CASRN 115-96-8, Supplemental File: Data Extraction and Data Evaluation Tables for Physical
and Chemical Property Studies. Washington, DC: Office of Chemical Safety and Pollution
Prevention.
U.S. EPA. (2021a). Draft systematic review protocol supporting TSCA risk evaluations for chemical
substances, Version 1.0: A generic TSCA systematic review protocol with chemical-specific
methodologies. (EPA Document #EPA-D-20-031). Washington, DC: Office of Chemical Safety
and Pollution Prevention, https://www.regulations. gov/document/EPA-HQ-QPPT-2021 -0414-
0005
U.S. EPA. (2021b). Final scope of the risk evaluation for di-isodecyl phthalate (DIDP) (1,2-
benzenedicarboxylic acid, 1,2-diisodecyl ester and 1,2-benzenedicarboxylic acid, di-C9-11-
branched alkyl esters, ClO-rich); CASRN 26761-40-0 and 68515-49-1 [EPAReport], (EPA-740-
R-21-001). Washington, DC: Office of Chemical Safety and Pollution Prevention.
https://www.epa.gOv/svstem/files/documents/2021-08/casrn-26761-40-0-di-isodecvl-phthalate-
final-scope.pdf
U.S. EPA. (2021c). Final scope of the risk evaluation for di-isononyl phthalate (DINP) (1,2-benzene-
dicarboxylic acid, 1,2-diisononyl ester, and 1,2-benzenedicarboxylic acid, di-C8-10-branched
alkyl esters, C9-rich); CASRNs 28553-12-0 and 68515-48-0 [EPA Report], (EPA-740-R-21-
002). Washington, DC: Office of Chemical Safety and Pollution Prevention.
https://www.epa.gov/svstem/files/documents/2021-08/casrn-28553-12-0-di-isononyl-phthalate-
final-scope.pdf
Page 429 of 638
-------
U.S. EPA. (2022b). Draft TSCA screening level approach for assessing ambient air and water exposures
to fenceline communities (version 1.0) [EPA Report], (EPA-744-D-22-001). Washington, DC:
Office of Chemical Safety and Pollution Prevention, U.S. Environmental Protection Agency.
https://www.epa.gov/svstem/files/documents/2022-01/draft-fenceline-report sacc.pdf
U.S. EPA. (2023). Consumer Exposure Model (CEM) Version 3.2 User's Guide. Washington, DC.
https://www.epa.gov/tsca-screening-tools/consumer-exposure-model-cem-version-32-users-
guide
U.S. EPA. (2024a). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental File
Folder: Supplemental Information on Human Milk PBPK Verner Modeling Results. Washington,
DC: Office of Pollution Prevention and Toxics, Office of Chemical Safety and Pollution
Prevention.
U.S. EPA. (2024b). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental File
Folder: Consumer Modeling Risk Calculations, Sensitivity Analysis, and Visualizations of
Results and Environmental Water Quality Portal Data. Washington, DC: Office of Pollution
Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.
U.S. EPA. (2024c). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental
Information File: Benchmark Dose Modeling Results for TCEP. Washington, DC: Office of
Pollution Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.
U.S. EPA. (2024d). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental
Information File: Biosolids Screening Tool (BST) Modeling Results. Washington, DC: Office of
Pollution Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.
U.S. EPA. (2024e). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental
Information File: Consumer Exposure Modeling Inputs. Washington, DC: Office of Pollution
Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.
U.S. EPA. (2024f). Risk Evaluation for Tris(2-Chloroethyl) Phosphate (TCEP) - Supplemental
Information File: Consumer Exposure Modeling Results. Washington, DC: Office of Pollution
Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.
https://www.epa.gov/svstem/files/documents/2023-
12/tcep draft risk evaluation 20231207 hero public-release.pdf
U.S. EPA. (2024g). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental
Information File: E-FAST Modeling Results. Washington, DC: Office of Pollution Prevention
and Toxics, Office of Chemical Safety and Pollution Prevention.
U.S. EPA. (2024h). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental
Information File: Environmental Monitoring and Biomonitoring Concentrations Summary Table.
Washington, DC: Office of Pollution Prevention and Toxics, Office of Chemical Safety and
Pollution Prevention.
U.S. EPA. (2024i). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental
Information File: Environmental Monitoring Concentrations Reported by Media Type.
Washington, DC: Office of Pollution Prevention and Toxics, Office of Chemical Safety and
Pollution Prevention.
U.S. EPA. (2024j). Risk Evaluation for Tris(2-Chloroethyl) Phosphate (TCEP) - Supplemental
Information File: Exposure Air Concentration Risk Calculations. Washington, DC: Office of
Pollution Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.
U.S. EPA. (2024k). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental
Information File: Human Health Hazard Points of Departure Comparison Tables. Washington,
DC: Office of Pollution Prevention and Toxics, Office of Chemical Safety and Pollution
Prevention.
Page 430 of 638
-------
U.S. EPA. (20241). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental
Information File: IIOAC Modeling Input and Results. Washington, DC: Office of Pollution
Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.
U.S. EPA. (2024m). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental
Information File: Risk Calculator for Occupational Exposures. Washington, DC: Office of
Pollution Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.
U.S. EPA. (2024n). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental
Information File: Supplemental Information on Environmental Release and Occupational
Exposure Assessment. Washington, DC: Office of Pollution Prevention and Toxics, Office of
Chemical Safety and Pollution Prevention.
U.S. EPA. (2024o). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental
Information File: Water Quality Portal Processed Water Data. Washington, DC: Office of
Pollution Prevention and Toxics, Office of Chemical Safety and Pollution Prevention.
U.S. EPA. (2024p). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review
Protocol. Washington, DC: Office of Pollution Prevention and Toxics, Office of Chemical Safety
and Pollution Prevention.
U.S. EPA. (2024q). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review
Supplemental File: Data Extraction Information for Environmental Hazard and Human Health
Hazard Animal Toxicology and Epidemiology. Washington, DC: Office of Pollution Prevention
and Toxics, Office of Chemical Safety and Pollution Prevention.
https://www.regulations.gov/docket/EPA-HQ-OPPT-2018-0476
U.S. EPA. (2024r). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review
Supplemental File: Data Extraction Information for General Population, Consumer, and
Environmental Exposure. Washington, DC: Office of Pollution Prevention and Toxics, Office of
Chemical Safety and Pollution Prevention.
U.S. EPA. (2024s). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review
Supplemental File: Data Quality Evaluation and Data Extraction Information for Dermal
Absorption. Washington, DC: Office of Pollution Prevention and Toxics, Office of Chemical
Safety and Pollution Prevention.
U.S. EPA. (2024t). Risk Evaluation for Tris(2-Chloroethyl) Phosphate (TCEP) - Systematic Review
Supplemental File: Data Quality Evaluation and Data Extraction Information for Environmental
Fate and Transport. Washington, DC: Office of Pollution Prevention and Toxics, Office of
Chemical Safety and Pollution Prevention. https://www.regulations.gov/document/EPA-HQ-
OPPT-2023-0265-0029
U.S. EPA. (2024u). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review
Supplemental File: Data Quality Evaluation and Data Extraction Information for Environmental
Release and Occupational Exposure. Washington, DC: Office of Pollution Prevention and
Toxics, Office of Chemical Safety and Pollution Prevention.
U.S. EPA. (2024v). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review
Supplemental File: Data Quality Evaluation and Data Extraction Information for Physical and
Chemical Properties. Washington, DC: Office of Pollution Prevention and Toxics, Office of
Chemical Safety and Pollution Prevention.
U.S. EPA. (2024w). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review
Supplemental File: Data Quality Evaluation Information for Environmental Hazard. Washington,
DC: Office of Pollution Prevention and Toxics, Office of Chemical Safety and Pollution
Prevention.
U.S. EPA. (2024x). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review
Supplemental File: Data Quality Evaluation Information for General Population, Consumer, and
Page 431 of 638
-------
Environmental Exposure. Washington, DC: Office of Pollution Prevention and Toxics, Office of
Chemical Safety and Pollution Prevention.
U.S. EPA. (2024y). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review
Supplemental File: Data Quality Evaluation Information for Human Health Hazard Animal
Toxicology. Washington, DC: Office of Pollution Prevention and Toxics, Office of Chemical
Safety and Pollution Prevention. https://www.regulations.gov/docket/EPA-HQ-OPPT-2018-0476
U.S. EPA. (2024z). Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review
Supplemental File: Data Quality Evaluation Information for Human Health Hazard
Epidemiology. Washington, DC: Office of Pollution Prevention and Toxics, Office of Chemical
Safety and Pollution Prevention.
Umezu. T; Yonemoto. J: Soma. Y; Suzuki. T. (1998). Tris(2-chloroethyl) phosphate increases
ambulatory activity in mice: Pharmacological analyses of its neurochemical mechanism. Toxicol
Appl Pharmacol 148: 109-116. http://dx.doi.org/10.1006/taap.1997.8312
Van den Eede. N: Maho. W: Erratico. C: Neels. H; Covaci. A. (2013). First insights in the metabolism
of phosphate flame retardants and plasticizers using human liver fractions. Toxicol Lett 223: 9-
15. http://dx.doi.Org/10.1016/i.toxlet.2013.08.012
van der Veen. I; de Boer. J. (2012). Phosphorus flame retardants: properties, production, environmental
occurrence, toxicity and analysis [Review], Chemosphere 88: 1119-1153.
http://dx.doi.Org/10.1016/i.chemosphere.2012.03.067
Venier. M; Dove. A: Romanak. K; Backus. S: Hites. R. (2014). Flame retardants and legacy chemicals
in Great Lakes' water. Environ Sci Technol 48: 9563-9572. http://dx.doi.org/10.1021/es501509r
Verbruggen. EMJ; Rila. JP; Traas. TP: Posthuma-Doodeman. CJAM; Posthumus. R. (2005).
Environmental risk limits for several phosphate esters, with possible application as flame
retardant. (RIVM report 601501024/2005). Bilthoven, The Netherlands: Netherlands National
Institute for Public Health and the Environment.
https://rivm.openrepositorv.com/bitstream/handle/10029/255802/601501024.pdf?sequence=3&is
Allowed=v
Verner. MA: Ayotte. P; Muckle. G: Charbonneau. M; Haddad. S. (2009). A physiologically based
pharmacokinetic model for the assessment of infant exposure to persistent organic pollutants in
epidemiologic studies. Environ Health Perspect 117: 481-487.
http://dx.doi.org/10.1289/ehp.0800Q47
Verner. MA: Charbonneau. M; Lopez-Carrillo. L; Haddad. S. (2008). Physiologically based
pharmacokinetic modeling of persistent organic pollutants for lifetime exposure assessment: A
new tool in breast cancer epidemiologic studies. Environ Health Perspect 116: 886-892.
http://dx.doi.org/10.1289/ehp.10917
Wallner. P; Kundi. M; Moshammer. H; Piegler. K; Hohenblum. P; Scharf. S: Frohlich. M; Damberger.
B; Tapplere. P; Hutter. HP. (2012). Indoor air in schools and lung function of Austrian school
children. J Environ Monit 14: 1976-1982. http://dx.doi.org/10.1039/c2em30059a
Wambaugh. JF; Wetmore. BA; Ring. CL; Nicolas. C: Pearce. RG: Honda. GS: Dinallo. R; Angus. D;
Gilbert. J: Sierra. T; Badrinaravanan. A: Snodgrass. B; Brockman. A: Strock. C: Setzer. RW:
Thomas. RS. (2019). Assessing toxicokinetic uncertainty and variability in risk prioritization.
Toxicol Sci 172: 235-251. http://dx.doi.org/10.1093/toxsci/kfz205
Wang. C: Wang. P; Zhao. J: Fu. M; Zhang. L; Li. Y; Yang. R; Zhu. Y; Fu. J: Zhang. O: Jiang. G.
(2020a). Atmospheric organophosphate esters in the Western Antarctic Peninsula over 2014-
2018: Occurrence, temporal trend and source implication. Environ Pollut 267: 115428.
http: //dx. doi. or g/10.1016/i. envpol .2020.115428
Wang. D; Wang. P; Wang. Y; Zhang. W: Zhu. C: Sun. H; Matsiko. J: Zhu. Y; Li. Y; Meng. W: Zhang.
O: Jiang. G. (2019a). Temporal variations of PM2.5-bound organophosphate flame retardants in
Page 432 of 638
-------
different microenvironments in Beijing, China, and implications for human exposure. Sci Total
Environ 666: 226-234. http://dx.doi.org/10.1016/i.scitotenv.2019.02.076
Wang. G; Shi. H; Du. Z; Chen. H; Peng. J; Gao. S. (2017a). Bioaccumulation mechanism of
organophosphate esters in adult zebrafish (Danio rerio). Environ Pollut 229: 177-187.
http://dx.doi.Org/10.1016/i.envpol.2017.05.075
Wang. H; Jing. C: Peng. H; Liu. S: Zhao. H; Zhang. W: Chen. X: Hu. F. (2022). Parental whole life-
cycle exposure to tris (2-chloroethyl) phosphate (TCEP) disrupts embryonic development and
thyroid system in zebrafish offspring. Ecotoxicol Environ Saf 248: 114313.
http: //dx. doi. or g/10.1016/i. ecoenv .2022.114313
Wang. Q: Zhao. H; Wang. Y; Xie. Q: Chen. J: Quan. X. (2017b). Determination and prediction of
octanol-air partition coefficients for organophosphate flame retardants. Ecotoxicol Environ Saf
145: 283-288. http://dx.doi.Org/10.1016/i.ecoenv.2017.07.040
Wang. RY; Needham. LL. (2007). Environmental chemicals: from the environment to food, to breast
milk, to the infant [Review], J Toxicol Environ Health B Crit Rev 10: 597-609.
http://dx.doi.org/10.1080/109374007Q1389891
Wang. X: Chen. P; Zhao. L; Zhu. L; Wu. F. (2021). Transplacental Behaviors of Organophosphate Tri-
and Diesters Based on Paired Human Maternal and Cord Whole Blood: Efficiencies and Impact
Factors. Environ Sci Technol 55: 3091-3100. http://dx.doi.org/10.1021/acs.est.0cQ6095
Wang. X: Zhong. W: Xiao. B; Liu. Q: Yang. L; Covaci. A: Zhu. L. (2019b). Bioavailability and
biomagnification of organophosphate esters in the food web of Taihu Lake, China: Impacts of
chemical properties and metabolism. Environ Int 125: 25-32.
http://dx.doi.Org/10.1016/i.envint.2019.01.018
Wang. X: Zhu. L; Zhong. W: Yang. L. (2018a). Partition and source identification of organophosphate
esters in the water and sediment of Taihu Lake, China. J Hazard Mater 360: 43-50.
http://dx.doi.Org/10.1016/i.ihazmat.2018.07.082
Wang. Y; Kannan. P; Halden. RU; Kannan. K. (2019c). A nationwide survey of 31 organophosphate
esters in sewage sludge from the United States. Sci Total Environ 655: 446-453.
http: //dx. doi. or g/10.1016/i. scitotenv .2018.11.224
Wang. Y; Li. W: Martinez-Moral. MP: Sun. H; Kannan. K. (2019d). Metabolites of organophosphate
esters in urine from the United States: Concentrations, temporal variability, and exposure
assessment. Environ Int 122: 213-221. http://dx.doi.Org/10.1016/i.envint.2018.ll.007
Wang. Y; Li. Z; Tan. F; Xu. Y; Zhao. H; Chen. J. (2020b). Occurrence and air-soil exchange of
organophosphate flame retardants in the air and soil of Dalian, China. Environ Pollut 265 Pt. A:
114850. http://dx.doi.Org/10.1016/i.envpol.2020.l 14850
Wang. Y; Wu. X: Zhang. O: Zhao. H; Hou. M; Xie. O: Chen. J. (2018b). Occurrence, distribution, and
air-water exchange of organophosphorus flame retardants in a typical coastal area of China.
Chemosphere 211: 33 5-344. http://dx.doi.org/10.1016/i.chemosphere.2018.07.062
Watts. MJ: Linden. KG. (2008). Photooxidation and subsequent biodegradability of recalcitrant tri-alkyl
phosphates TCEP and TBP in water. Water Res 42: 4949-4954.
http://dx.doi.Org/10.1016/i.watres.2008.09.020
Watts. MJ: Linden. KG. (2009). Advanced oxidation kinetics of aqueous trialkyl phosphate flame
retardants and plasticizers. Environ Sci Technol 43: 2937-2942.
http://dx.doi.org/10.1021/es8031659
Wegmann. F; Cavin. L; Macleod. M; Scheringer. M; Hungerbuhler. K. (2009). The OECD software tool
for screening chemicals for persistence and long-range transport potential. Environ Modell Softw
24: 228-237. http://dx.doi.Org/10.1016/i.envsoft.2008.06.014
Wei. GL: Li. DO: Zhuo. MN: Liao. YS: Xie. ZY: Guo. TL: Li. JJ; Zhang. SY: Liang. ZO. (2014).
Organophosphorus flame retardants and plasticizers: Sources, occurrence, toxicity and human
exposure [Review], Environ Pollut 196C: 29-46. http://dx.doi.Org/10.1016/i.envpol.2014.09.012
Page 433 of 638
-------
Weil. ED. (2001). Kirk-Othmer Encyclopedia of Chemical Technology
Flame Retardants, Phosphorus. New York, NY: John Wiley & Sons.
http://dx.doi.Org/https://onlinelibrary.wilev.com/doi/abs/10.1002/0471238961.160815192305091
2.a01.pub2
Welke. B; Ettlinger. K; Riederer. M. (1998). Sorption of volatile organic chemicals in plant surfaces.
Environ Sci Technol 32: 1099-1104.
Weschler. CJ: Nazaroff WW. (2010). SVOC partitioning between the gas phase and settled dust indoors
[Review], Atmos Environ 44: 3609-3620. http://dx.doi.Org/10.1016/i.atmosenv.2010.06.029
Weselv. ML: Doskev. PV: Shannon. JD. (2002). Deposition parameterizations for the Industrial Source
Complex (ISC3) model. (ANL/ER/TR-01/003). Argonne, IL: Argonne National Lab.
Westat. (1987). Household solvent products: A national usage survey [EPA Report], (EPA-OTS 560/5-
87-005). Washington, DC: Office of Toxic Substances, Office of Pesticides and Toxic
Substances. https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockev=P100754Q.txt
Westerhoff. P; Yoon. Y; Snyder. S: Wert. E. (2005). Fate of endocrine-disruptor, pharmaceutical, and
personal care product chemicals during simulated drinking water treatment processes. Environ
Sci Technol 39: 6649-6663. http://dx.doi.org/10.1021/es0484799
White. PR: Widdowson. EM: Woodard. HQ: Dickerson. JWT. (1991). The composition of body tissues
(II) Fetus to young adult. Br J Radiol 64: 149-159.
Whittaker. C: Rice. F; McKernan. L; Dankovic. D; Lentz. T; Macmahon. K; Kuempel. E; Zumwalde. R;
Schulte. P. (2016). Current Intelligence Bulletin 68: NIOSH Chemical Carcinogen Policy. US
Department of Health and Human Services.
https://ntrl.ntis. gov/NTRL/dashboard/ searchResults/titleDetail/PB2017101413 .xhtml
Willming. MM: Lilavois. CR; Barron. MG: Raimondo. S. (2016). Acute toxicity prediction to
threatened and endangered species using Interspecies Correlation Estimation (ICE) models.
Environ Sci Technol 50: 10700-10707. http://dx.doi.org/10.1021/acs.est.6b03009
Wolschke. H; Suehring. R; Mi. W: Moeller. A: Xie. Z; Ebinghaus. R. (2016). Atmospheric occurrence
and fate of organophosphorus flame retardants and plasticizer at the German coast. Atmos
Environ 137: 1-5. http://dx.doi.Org/10.1016/i.atmosenv.2016.04.028
Wong. F; De Wit. CA; Newton. SR. (2018). Concentrations and variability of organophosphate esters,
halogenated flame retardants, and polybrominated diphenyl ethers in indoor and outdoor air in
Stockholm, Sweden. Environ Pollut 240: 514-522.
http://dx.doi.Org/10.1016/i.envpol.2018.04.086
Woudneh. MB: Benskin. JP; Wang. G: Grace. R; Hamilton. MC: Cosgrove. JR. (2015). Quantitative
determination of 13 organophosphorous flame retardants and plasticizers in a wastewater
treatment system by high performance liquid chromatography tandem mass spectrometry. J
Chromatogr A 1400: 149-155. http://dx.doi.Org/10.1016/i.chroma.2015.04.026
WSDE. (2022). Chemicals of emerging concern in pretreated industrial wastewater in Northwestern
Washington state: Screening study results, 2021. Olympia, WA.
https://apps.ecologv.wa.gov/publications/SummaryPages/2203013.html
Wu. Y; Venier. M; Salamova. A. (2020). Spatioseasonal variations and partitioning behavior of
organophosphate esters in the Great Lakes atmosphere. Environ Sci Technol 54: 5400-5408.
http://dx.doi.org/10.1021/acs.est.9b07755
Xie. Z; Wang. Z; Magand. O: Thollot. A: Ebinghaus. R; Mi. W: Dommergue. A. (2020). Occurrence of
legacy and emerging organic contaminants in snow at Dome C in the Antarctic. Sci Total
Environ 741: 140200. http://dx.doi.org/10.1016/i.scitotenv.2020.140200
Xing. L; Zhang. O: Sun. X: Zhu. H; Zhang. S: Xu. H. (2018). Occurrence, distribution and risk
assessment of organophosphate esters in surface water and sediment from a shallow freshwater
Lake, China. Sci Total Environ 636: 632-640. http://dx.doi.org/10.1016/i.scitotenv.2018.04.320
Page 434 of 638
-------
Xu. F; Giovanoulis. G; van Waes. S; Padilla-Sanchez. JA; Papadopoulou. E; Magner. J; Haug. LS;
Neels. H; Covaci. A. (2016). Comprehensive study of human external exposure to
organophosphate flame retardants via air, dust, and hand wipes: The importance of sampling and
assessment strategy. Environ Sci Technol 50: 7752-7760.
http://dx.doi.org/10.1021/acs.est.6b00246
Xu. T; Li. P; Wu, S: Lei. L; He. D. (2017). Tris(2-chloroethyl) phosphate (TCEP) and tris(2-
chloropropyl) phosphate (TCPP) induce locomotor deficits and dopaminergic degeneration in
Caenorhabditis elegans. Toxicology Research 6: 63-72. http://dx.doi.org/10.1039/c6txQ0306k
Yadav. IC: Devi. NL; Li. J: Zhang. G. (2018). Organophosphate ester flame retardants in Nepalese soil:
Spatial distribution, source apportionment and air-soil exchange assessment. Chemosphere 190:
114-123. http://dx.doi.org/10.1016/i.chemosphere.2017.09.112
Yaman. B; Dumanoglu. Y; Odabasi. M. (2020). Measurement and modeling the phase partitioning of
organophosphate esters using their temperature-dependent octanol-air partition coefficients and
vapor pressures. Environ Sci Technol 54: 8133-8143. http://dx.doi.org/10.1021/acs.est.0c02823
Yang. F; Ding. J: Huang. W: Xie. W: Liu. W. (2014). Particle size-specific distributions and preliminary
exposure assessments of organophosphate flame retardants in office air particulate matter.
Environ Sci Technol 48: 63-70. http://dx.doi.org/10.1021/es403186z
Yang. S: Wu. J: Wang. H; Yang. O; Zhang. H; Yang. L; Li. D; Deng. Y; Zhong. Y; Peng. P. (2023).
New dechlorination products and mechanisms of tris(2-chloroethyl) phosphate by an anaerobic
enrichment culture from a vehicle dismantling site. Environ Pollut 338: 122704.
http: //dx. doi. or g/10.1016/i. envpol .2023.122704
Yang. W: Zhao. F; Fang. Y; Li. L; Li. C: Ta. N. (2018a). lH-nuclear magnetic resonance metabolomics
revealing the intrinsic relationships between neurochemical alterations and neurobehavioral and
neuropathological abnormalities in rats exposed to tris(2-chloroethyl)phosphate. Chemosphere
200: 649-659. http://dx.doi.org/10.1016/i.chemosphere.2018.02.056
Yang. WL; Braun. JM; Vuong. AM: Percy. Z; Xu. YY; Xie. CC: Deka. R; Calafat. AM: Ospina. M;
Burris. HH; Yolton. K; Cecil. KM: Lanphear. BP: Chen. AM. (2022). Associations of gestational
exposure to organophosphate esters with gestational age and neonatal anthropometric measures:
The HOME study*. Environ Pollut 316: 120516. http://dx.doi.org/10.1016/i.envpol.2022.120516
Yang. X: Fan. D; Gu. W: Liu. J: Shi. L; Zhang. Z; Zhou. L; Ji. G. (2021). Aerobic and anaerobic
biodegradability of organophosphates in activated sludge derived from kitchen garbage biomass
and agricultural residues. Frontiers in Bioengineering and Biotechnology 9: 649049.
http://dx.doi.org/10.3389/fbioe.2021.649049
Yang. X: Sun. L; Xiang. J: Hu. S: Su. S. (2013). Pyrolysis and dehalogenation of plastics from waste
electrical and electronic equipment (WEEE): A review. Waste Manag 33: 462-473.
http://dx.doi.Org/10.1016/i.wasman.2012.07.025
Yang. Y; Xiao. Y; Chang. Y; Cui. Y; Klobucar. G: Li. M. (2018b). Intestinal damage, neurotoxicity and
biochemical responses caused by tris (2-chloroethyl) phosphate and tricresyl phosphate on
earthworm. Ecotoxicol Environ Saf 158: 78-86. http://dx.doi.Org/10.1016/i.ecoenv.2018.04.012
Yasuhara. A. (1995). Chemical components in leachates from hazardous wastes landfills in Japan.
Toxicol Environ Chem 51: 113-120. http://dx.doi.org/10.1080/027722495Q9358229
Yasuhara. A: Shiraishi. H; Nishikawa. M; Yamamoto. T; Nakasugi. O: Okumura. T; Kenmotsu. K;
Fukui. H; Nagase. M; Kawagoshi. Y. (1999). Organic components in leachates from hazardous
waste disposal sites. Waste Manag Res 17: 186-197. http://dx.doi.org/10.1034/i. 1399-
3070.1999.00038.x
Yonemoto. J: Shiraishi. H; Soma. Y; Inaba. K; Sone. H; Kobavashi. S. (1997). Use of rat embryo limb
bud cell cultures to screen organochlorine compounds detected in the water and sediment of
rivers in Tokyo metropolis for developmental toxicity. Toxicol Environ Chem 62: 125-133.
http://dx.doi.org/10.1080/027722497093585Q2
Page 435 of 638
-------
Zhang. H; Zhou. 0; Xie. Z; Zhou. Y; Tu. C; Fu. C; Mi. W; Ebinghaus. R; Christie. P; Luo. Y. (2018a).
Occurrences of organophosphorus esters and phthalates in the microplastics from the coastal
beaches in north China. Sci Total Environ 616-617: 1505-1512.
http: //dx. doi. or g/10.1016/i. scitotenv .2017.10.163
Zhang. P; Meng. F; Xia. Y; Leng. Y; Cui. J. (2024). Deriving seawater quality criteria of tris(2-
chloroethyl) phosphate for ecological risk assessment in China seas through species sensitivity
distributions. J Environ Manage 349: 119482. http://dx.doi.Org/10.1016/i.ienvman.2023.l 19482
Zhang. S: Gitungo. S: Axe. L; Dyksen. JE; Raczko. RF. (2016a). A pilot plant study using conventional
and advanced water treatment processes: Evaluating removal efficiency of indicator compounds
representative of pharmaceuticals and personal care products. Water Res 105: 85-96.
http://dx.doi.Org/10.1016/i.watres.2016.08.033
Zhang. S: Rosati. A: Harrison. MJ. (2009). Detection of multidecadal oceanic variability by ocean data
assimilation in the context of a "perfect" coupled model. J Geophys Res Oceans 114.
http://dx.doi.org/10.1029/2008JC0Q5261
Zhang. W: Guo. C: Lv. J: Li. X: Xu. J. (2021). Organophosphate esters in sediment from Taihu Lake,
China: Bridging the gap between riverine sources and lake sinks. Front Env Sci Eng 16: 30.
http://dx.doi.org/10.1007/sll783-021-1464-9
Zhang. W: Zhang. Y: Hou. J: Xu. T: Yin. W: Xiong. W: Lu. W: Zheng. H: Chen. J: Yuan. J. (2017a).
Tris (2-chloroethyl) phosphate induces senescence-like phenotype of hepatocytes via the
p21Wafl/Cipl-Rb pathway in a p53-independent manner. Environ Toxicol Pharmacol 56: 68-75.
http://dx.doi.Org/10.1016/i.etap.2017.08.028
Zhang. W: Zhang. Y; Wang. Z; Xu. T; Huang. C: Yin. W: Wang. J: Xiong. W: Lu. W: Zheng. Ft; Yuan.
L (2016b). Tris(2-chloroethyl)phosphate-induced cell growth arrest via attenuation of SIRT1-
independent PI3K/Akt/mTOR pathway. J Appl Toxicol 36: 914-924.
http: //dx. doi. or g/10.1002/i at. 3223
Zhang. W: Zhang. Y: Xu. T: Wang. Z: Wang. J: Xiong. W: Lu. W: Zheng. H: Yuan. J. (2016c).
Involvement of ROS-mediated mitochondrial dysfunction and SIRT3 down-regulation in tris(2-
chloroethyl) phosphate-induced cell cycle arrest. Toxicology Research 5: 461-470.
http://dx.doi.org/10.1039/c5tx00229i
Zhang. Y; Zhang. W: Hou. J: Wang. X: Zheng. H; Xiong. We: Yuan. J. (2017b). Combined effect of
tris(2-chloroethyl) phosphate and benzo (a) pyrene on the release of IL-6 and IL-8 from HepG2
cells via the EGFR-ERK1/2 signaling pathway. RSC Adv 7: 54281-54290.
http://dx.doi.org/10.1039/c7rall273d
Zhang. Y; Zheng. X: Wei. L; Sun. R; Guo. H; Liu. X: Liu. S: Li. Y; Mai. B. (2018b). The distribution
and accumulation of phosphate flame retardants (PFRs) in water environment. Sci Total Environ
630: 164-170. http://dx.doi.org/10.1016/i.scitotenv.2018.02.215
Zhang. Z; Xu. Y; Wang. Y; Li. Z; Yang. C: Rodgers. TFM; Tan. F. (2022). Occurrence and distribution
of organophosphate flame retardants in the typical soil profiles of the Tibetan Plateau, China. Sci
Total Environ 807: 150519. http://dx.doi.org/10.1016/i.scitotenv.2021.150519
Zhao. F; Chen. M; Gao. F; Shen. H; Hu. J. (2017). Organophosphorus flame retardants in pregnant
women and their transfer to chorionic villi. Environ Sci Technol 51: 6489-6497.
http://dx.doi.org/10.1021/acs.est.7b01122
Zhao. H; Zhao. F; Liu. J: Zhang. S: Mu. D; An. L; Wan. Y; Hu. J. (2018). Trophic transfer of
organophosphorus flame retardants in a lake food web. Environ Pollut 242: 1887-1893.
http://dx.doi.Org/10.1016/i.envpol.2018.07.077
Zhao. Y; Yin. L; Dong. F; Zhang. W: Hu. F. (2021). Effects of tris (2-chloroethyl) phosphate (TCEP) on
survival, growth, histological changes and gene expressions in juvenile yellow catfish
Pelteobagrus fulvidraco. Environ Toxicol Pharmacol 87: 103699.
http ://dx. doi. org/10.1016/i. etap .2021.103699
Page 436 of 638
-------
Zhong. M; Wu, H; Mi. W; Li. F; Ji. C; Ebinghaus. R; Tang. J; Xie. Z. (2018). Occurrences and
distribution characteristics of organophosphate ester flame retardants and plasticizers in the
sediments of the Bohai and Yellow Seas, China. Sci Total Environ 615: 1305-1311.
http: //dx. doi. or g/10.1016/i. scitotenv .2017.09.272
Zhu. H; Zhang. H; Lu. K; Yang. S; Tang. X; Zhou. M; Sun. G; Zhang. Z; Chu. H. (2022). Chlorinated
organophosphate flame retardants impair the lung function via the IL-6/JAK/STAT signaling
pathway. Environ Sci Technol 56: 17858-17869. http://dx.doi.org/10.1021/acs.est.2c05357
Zhu. M; He. L; Liu. J: Long. Y; Shentu. J: Lu. L; Shen. D. (2023). Dynamic processes in conjunction
with microbial response to unveil the attenuation mechanisms of tris (2-chloroethyl) phosphate
(TCEP) in non-sanitary landfill soils. Environ Pollut 316: 120666.
http ://dx. doi. org/10.1016/i. envpol .2022.120666
Page 437 of 638
-------
APPENDICES
Appendix A ABBREVIATIONS, ACRONYMS, AND GLOSSARY OF
SELECT TERMS
A.l Key Abbreviations and Acronyms
AC
Acute exposure concentrations
AChE
Acetyl cholinesterase
ADC
Average daily concentrations
ADME
Absorption, distribution, metabolism, and elimination
AERMOD
American Meteorological Society (AMS)/EPA Regulatory Model
AF
Assessment factor
ALP
Alkaline phosphatase
ALT
Alanine transferase
AST
Aspartate transaminase
ATSDR
Agency for Toxic Substances and Disease Registry
BAF
Bioaccumulation factor
BCCP
Bis(2-chloroethyl) carboxymethyl phosphate
BCEP
Bis(2-chloroethyl) phosphate
BCF
Bioconcentration factor
BCGP
The glucuronide of bis(2-chloroethyl) 2-hydroxyethyl phosphate
BCHP
Bis(2-chloroethyl) hydrogen phosphate
BLS
Bureau of Labor Statistics
BMD
Benchmark dose
BMDL
Benchmark dose lower confidence limit
BMF
Biomagnification factor
BMR
Benchmark response
BSAF
Biota-sediment accumulation factor
CASRN
Chemical Abstracts Service Registry Number
CBI
Confidential business information
CDR
Chemical Data Reporting (Rule)
CEPA
Canadian List of Toxic Substances
CERCLA
Comprehensive Environmental Response, Compensation and Liability Act
CFR
Code of Federal Regulations
ChV
Chronic health value
CI
Confidence interval
coc
Concentration(s) of concern
cou
Condition of use
CoCAP
Cooperative Chemicals Assessment Program
CPSA
Consumer Product Safety Act
CPSC
Consumer Product Safety Commission
CSCL
Chemical Substances Control Law
CSF
Cancer slope factor
CSHO
Certified Safety and Health Official
CTD
Characteristic travel distance
DIY
Do-it-yourself
DMR
Discharge Monitoring Report
Page 438 of 638
-------
DOT
Department of Transportation
DRAS
(Hazardous Waste) Delisting Risk Assessment Software (EPA model)
DT50
Time needed to eliminate 50 percent of the substance
DT90
Time needed to eliminate 90 percent of the substance
DWTP
Drinking water treatment plant
EC50
Effect concentration at which 50 percent of test organisms exhibit an effect
ECHA
European Chemicals Agency
ECOSAR
Ecological Structure Activity Relationships (model)
EPA
Environmental Protection Agency
EPCRA
Emergency Planning and Community Right-to-Know Act
ESD
Emission Scenario Document
EU
European Union
FIR
Food intake rate
GS
Generic Scenario
HC05
Hazard concentration that is protective of 95 percent of the species in the sensitivity
distribution
HEC
Human equivalent concentration
HED
Human equivalent dose
HERO
Health and Environmental Research Online (Database)
HHE
Health hazard evaluation
IARC
International Agency for Research on Cancer
IMAP
Inventory Multi-Tiered Assessment and Prioritization
IR
Ingestion rate
IRIS
Integrated Risk Information System
IUR
Inhalation unit risk
Kaw
Ainwater partition coefficient
Koc
Soil organic carbon:water partitioning coefficient
Kow
Octanol:water partition coefficient
Kp
Permeability coefficient
LADC
Lifetime average daily concentrations
LADD
Lifetime average daily dose
LCD
Life cycle diagram
LC50
Lethal concentration at which 50 percent of test organisms die
LD50
Lethal dose at which 50 percent of test organisms die
LOAEL
Lowest-observable-adverse-effect level
LOD
Limit of detection
LOEC
Lowest-observed-effect concentration
LOQ
Limit of quantification
Log Kaw
Logarithmic ainwater partition coefficient
Log Koa
Logarithmic octanol:air partition coefficient
Log Koc
Logarithmic organic carbon:water partition coefficient
Log Kow
Logarithmic octanol:water partition coefficient
LRAT
Long-range transport via long-range atmospheric transport
MOA
Mode of action
MOE
Margin of exposure
MSW
Municipal solid waste
MSWLF
Municipal solid waste landfills
NAICS
North American Industry Classification System
NATA
National Scale Air-Toxics Assessment
Page 439 of 638
-------
ND
Non-detect
NEI
National Emissions Inventory
NHANES
National Health and Nutrition Examination Survey
NICNAS
National Industrial Chemicals Notification and Assessment Scheme
NIH
National Institutes of Health
NIOSH
National Institute for Occupational Safety and Health
NITE
National Institute of Technology and Evaluation
NMAM
NIOSH Manual of Analytical Methods
NO A A
National Oceanic and Atmospheric Administration
NOEL
No-observed-effect level
NOAEL
No-observed-adverse-effect level
NPDES
National Pollutant Discharge Elimination System
NTP
National Toxicology Program
NWIS
National Water Information System
OCSPP
Office of Chemical Safety and Pollution Prevention
OECD
Organisation for Economic Co-operation and Development
OES
Occupational exposure scenario
ONU
Occupational non-user
OPP
Office of Pesticide Programs
OPPT
Office of Pollution Prevention and Toxics
OSHA
Occupational Safety and Health Administration
PBPK
Physiologically based pharmacokinetic
PBZ
Personal breathing zone
PECO
Population, exposure, comparator, and outcome
PEL
Permissible exposure limit (OSHA)
PESS
Potentially exposed or susceptible subpopulations
PMOC
Persistent mobile organic compound
POD
Point of departure
POTW
Publicly owned treatment works
PPE
Personal protective equipment
PV
Production volume
QSAR
Quantitative structure-activity relationship (model)
RCRA
Resource Conservation and Recovery Act
REACH
Registration, Evaluation, Authorization and Restriction of Chemicals (European Union)
RP
Respirable particle
RQ
Risk quotient
SACC
Scientific Advisory Committee on Chemicals
SCADC
Subchronic average daily concentration
SCE
Sister chromatid exchange
SDS
Safety data sheet
SIDS
Screening Information Dataset
SOC
Standard Occupational Classification (BLS codes)
SSD
Species sensitivity distribution
STEL
Short-term exposure limit
STEV
Short-term occupational exposure value
STORET
STOrage and RETrieval and Water Quality exchange
SVOC
Semi-volatile compound
TE
Transfer efficiency
TESIE
Toddler's Exposure to SVOCs in the Indoor Environment (study)
Page 440 of 638
-------
TGD
Technical Guidance Document (European Commission)
TCEP
Tris(2-chloroethyl) phosphate
TMF
Trophic magnification factor
TRI
Toxics Release Inventory
TRV
Toxicity reference value
TSCA
Toxic Substances Control Act
TWA
Time-weighted average
UF
Uncertainty factor
U.S.
United States
USGS
United States Geological Survey
V6
2,2-Bis(chloromethyl)-propane-l,3-diyltetrakis(2-chloroethyl) bisphosphate
voc
Volatile organic compound
VP
Vapor pressure
Web-ICE
Web-based Interspecies Correlation Estimation
WHO
World Health Organization
WQP
Water Quality Portal
WWTP
Wastewater treatment plant
7Q10
The lowest 7-day average flow that occurs (on average) once every 10 years
30Q5
The lowest 30-day average flow that occurs (on average) once every 5 years
A.2 Glossary of Select Terms
Condition of use (COU) (15 U.S.C. § 2602(4)): "means the circumstances, as determined by the
Administrator, under which a chemical substance is intended, known, or reasonably foreseen to be
manufactured, processed, distributed in commerce, used, or disposed of."
Margin of exposure (MOE) (U.S. EPA. 2002a): "a numerical value that characterizes the amount of
safety to a toxic chemical-a ratio of a toxicological endpoint (usually a NOAEL [no observed adverse
effect level]) to exposure. The MOE is a measure of how closely the exposure comes to the NOAEL."
Mode of action (MOA) (U.S. EPA 2000c): "a series of key events and processes starting with
interaction of an agent with a cell and proceeding through operational and anatomical changes causing
disease formation."
Point of departure (POD) (U.S. EPA 2002a): "dose that can be considered to be in the range of
observed responses, without significant extrapolation. A POD can be a data point or an estimated point
that is derived from observed dose-response data. A POD is used to mark the beginning of extrapolation
to determine risk associated with lower environmentally relevant human exposures."
Potentially exposed or susceptible subpopulations (PESS) (15 U.S.C. § 2602(12)): "means a group of
individuals within the general population identified by the Agency who, due to either greater
susceptibility or greater exposure, may be at greater risk than the general population of adverse health
effects from exposure to a chemical substance or mixture, such as infants, children, pregnant women,
workers, or the elderly."
Reasonably available information (40 CFR 702.33): "means information that EPA possesses or can
reasonably generate, obtain, and synthesize for use in risk evaluations, considering the deadlines
specified in TSC A section 6(b)(4)(G) for completing such evaluation. Information that meets the terms
of the preceding sentence is reasonably available information whether or not the information is
confidential business information, that is protected from public disclosure under TSCA section 14."
Page 441 of 638
-------
Routes (40 CFR 702.33): "means the ways a chemical substance enters an organism after contact, e.g.,
by ingestion, inhalation, or dermal absorption."
Sentinel exposure (40 CFR 702.33): "means the exposure from a chemical substance that represents the
plausible upper bound of exposure relative to all other exposures within a broad category of similar or
related exposures."
Page 442 of 638
-------
Appendix B REGULATORY AND ASSESSMENT HISTORY
B.l Federal Laws and Regulations
Table Apx B-l. Federal Laws and Regulations
Statutes/Regulations
Description of Authority/Regulation
Description of Regulation
EPA statutes/regulations
Toxic Substances
Control Act (TSCA) -
Section 5(a)
Once EPA finalizes a Significant New Use
Rule (SNUR) determining that a use of a
chemical substance is a significant new use
under TSCA section 5(a), persons are
required to submit a significant new use
notice (SNUN) to EPA at least 90 days
before they manufacture (including import)
or process the chemical substance for that
use.
In June, 2023, EPA proposed a significant
new use rule (SNUR) to designate
manufacture or processing of TCEP for
any use as a significant new use, with the
exception that the conditions of use the
Agency expects to consider within the
scope of the TSCA section 6 risk
evaluations are not proposed as significant
new uses. (88 FR 40741. June 22. 2023).
Toxic Substances
Control Act (TSCA) -
Section 6(b)
EPA is directed to identify high-priority
chemical substances for risk evaluation;
and conduct risk evaluations on at least 20
high priority substances no later than three
and one-half years after the date of
enactment of the Frank R. Lautenberg
Chemical Safety for the 21st Century Act.
TCEP is one of the 20 chemicals EPA
designated as a High-Priority Substance for
risk evaluation under TSCA (84 FR 71924.
December 30, 2019). Designation of TCEP
as high-priority substance constitutes the
initiation of the risk evaluation on the
chemical.
Toxic Substances
Control Act (TSCA) -
Section 8(a)
The TSCA section 8(a) CDR Rule requires
manufacturers (including importers) to
give EPA basic exposure-related
information on the types, quantities and
uses of chemical substances produced
domestically and imported into the United
States.
TCEP manufacturing (including
importing), processing and use information
is reported under the CDR rule (40 CFR
part 711).
Toxic Substances
Control Act (TSCA) -
Section 8(b)
EPA must compile, keep current and
publish a list (the TSCA Inventory) of each
chemical substance manufactured
(including imported) or processed for
commercial purposes in the United States.
TCEP was on the initial TSCA Inventory
and therefore was not subject to EPA's
new chemicals review process under TSCA
section 5 (60 FR 16309. March 29. 1995).
The chemical is on the active inventory.
Toxic Substances
Control Act (TSCA) -
section 8(d)
Provides EPA with authority to issue rules
requiring manufacturers (including
importers), processors, and distributors of a
chemical substance or mixture to submit
lists and/or copies of ongoing and
completed, unpublished health and safety
studies. EPA's Health and Safety Data
Reporting Rule at 40 CFRpart 716
generally requires such submissions for
manufacturers (including importers) and (if
specified) processors of substances covered
by part 716.
Two submissions received in 2021 (U.S.
EPA, Chemical Data Access Tool,
accessed November 25, 2022).
Page 443 of 638
-------
Statutes/Regulations
Description of Authority/Regulation
Description of Regulation
Toxic Substances
Control Act (TSCA) -
Section 4
Provides EPA with authority to issue rules
and orders requiring manufacturers
(including importers) and processors to test
chemical substances and mixtures.
Three chemical data submissions from test
rules received for TCEP: all three were
monitoring reports (1978, 1980, and 1981)
(U.S. EPA. ChemView, accessed April 3.
2019).
Emergency Planning
and Community
Right-To-Know Act
(EPCRA) - Section
313
EPCRA Section 313 - also known as the
Toxic Release Inventory (TRI), requires
annual reporting from facilities in specific
industry sectors that employ 10 or more
full-time equivalent employees and that
manufacture, process or otherwise use a
TRI-listed chemical in quantities above
threshold levels. A facility that meets
reporting requirements must submit a
reporting form for each chemical for which
it triggered reporting, providing data across
a variety of categories, including activities
and uses of the chemical, releases, and
other waste management (e.g., quantities
recycled, treated, combusted) and pollution
prevention activities (under section 6607 of
the Pollution Prevention Act). These data
include on- and off-site data as well as
multimedia data (i.e.. air, land, and water).
TCEP is a listed substance subject to
reporting requirements under 40 CFR
372.65 effective as of November 30, 2022.
B.2 State Laws and Regulations
Table Apx B-2. State Laws and Regulations
State Actions
Description of Action
State Prohibitions
Three states have adopted prohibitions for the use of TCEP in children's products,
including Marvland (MD Health Gen § 24-306). New York (TRIS-free Children and
Babies Act (NY Envir Conser § 37-0701 et seq.)). Minnesota (Four flame Retardants in
Furniture Foam and Children's Products (Minn. Stat. § 325F.071)).
California adopted a prohibition, effective on January 1, 2020, on the selling and
distribution in commerce of new, not previously owned juvenile products, mattresses, or
upholstered furniture that contains, or a constituent component of which contains,
covered flame retardant chemicals at levels above 1.000 parts per million (A.B. 2998.
Legislative Council. Sess. 2017-2018. C.A. 2018).
State Drinking Water
Standards and
Guidelines
Minnesota developed a health-based guidance value for TCEP in drinking water (Minn
R. Chap. 4720).
Chemicals of High
Concern to Children
Several states have adopted reporting laws for chemicals in children's products
containing TCEP. including Maine (38 MRSA Chapter 16-D). Minnesota (Toxic Free
Kids Act Minn. Stat. 116.9401 to 116.9407). Oregon (Toxic-Free Kids Act. Senate Bill
478. 2015). Vermont (18 V.S.A § 1776) and Washington State (Wash. Admin. Code
173-334-130).
Other
California listed TCEP on Proposition 65 in 1992 due to cancer (Cal Code Regs. Title
27. §27001).
Page 444 of 638
-------
State Actions
Description of Action
California issued a Health Hazard Alert for TCEP (Hazard Evaluation Svstem and
Information Service. 2016).
California lists TCEP as a designated priority chemical for biomonitoring (California
SB 1379).
TCEP is listed as a Candidate Chemical under California's Safer Consumer Products
Program (Health and Safetv Code § 25252 and 25253). The regulation for Children's
Foam-Padded Sleeping Products containing TCEP as a Priority Product went into effect
on July 1, 2017: Manufacturers of this product must notify the Department by
September 1, 2017 (California Department of Toxic Substances Control, Accessed April
12, 2019).
B.3 International Laws and Regulations
Table Apx B-3. International Laws and Regulations
Country/
Organization
Requirements and Restrictions
Canada
TCEP (Ethanol, 2-chloro-, phosphate (3:1)) is on the Canadian List of Toxic
Substances (CEPA 1999 Schedule 1).
TCEP was added to Schedule 2 of the Canada Consumer Product Sa fety Act
(CCPSA), based on concerns for carcinogenicity and impaired fertility.
(Government Canada Chemical Safetv portal. Accessed April 10. 2019).
In January 2013. a Significant New Activitv was adopted for TCEP (Canada
Gazette, April 3. 2014; Vol. 148. No. 9).
European Union
In June 2017, TCEP was added to Annex XIV of REACH (Authorisation List) with
a sunset date of August 21. 2015 (European Chemicals Agencv (ECHA. 2019)
database. Accessed April 10. 2019).
In 2010, TCEP was listed on the Candidate list as a Substance of Very High
Concern (SVHC) under regulation (EC) No 1907/2006 - REACH (Registration.
Evaluation. Authorization and Restriction of Chemicals due to its reproductive
toxicitv (category 57C)).
Australia
Ethanol, 2-chloro-, phosphate (3:1) (TCEP) was assessed under Human Health Tier
II and III of the Inventory Multi-Tiered Assessment and Prioritisation (IMAP). Uses
reported include commercial: (NICNAS. 2016. Ethanol, 2-chloro-, phosphate (3:1):
Human health tier II assessment. Accessed April 8. 2019) (NICNAS. 2017.
Ethanol, 2-chloro-, phosphate (3:1): Human health tier III assessment. Accessed
April 8, 2019).
Japan
TCEP is regulated in Japan under the following legislation:
• Act on the Evaluation of Chemical Substances and Regulation of Their
Manufacture, etc. (Chemical Substances Control Law; CSCL),
• Act on Confirmation, etc. of Release Amounts of Specific Chemical Substances
in the Environment and Promotion of Improvements to the Management Thereof,
Page 445 of 638
-------
Country/
Organization
Requirements and Restrictions
• Air Pollution Control Law
(National Institute of Technology and Evaluation INITEI Chemical Risk
Information Platform ICHRIPI. April 8. 2019).
Basel Convention
Waste substances and articles containing or contaminated with polychlorinated
biphenyls (PCBs) and/or polychlorinated terphenyls (PCTs) and/or polybrominated
biphenyls (PBBs) are listed as a category of waste under the Basel Convention.
Although the United States is not currently a party to the Basel Convention, this
treaty still affects U.S. importers and exporters.
http://www.basel.int/Portals/4/Basel%20Convention/docs/text/BaselConventionText
-e.pdf.
B.4 Assessment History
Table Apx B-4. Assessment History of TCEP
Authoring Organization
Publication
EPA publications
U.S. EPA, Superfund Health Risk Technical Support
Center, Office of Research and Development (ORD)
Provisional Peer-Reviewed Toxicitv Values (PPRTV)
for Tris(2-chloroethvl)phosphate (TCEP) (CASRN 115-
96-8) U.S. EPA (2009)
U.S. EPA, Design for the Environment Program
Design for the Environment (DfE) Alternatives
Assessments
Other U.S.-based organizations
Agency for Toxic Substances and Disease Registry
(ATSDR)
Toxicological Profile for Phosphate Ester Flame
Retardants (2012)
National Toxicology Program (NTP), National
Institutes of Health (NIH)
Technical Report on Toxicology and Carcinogenesis
Studies of Tris(2-chloroethvl) Phosphate (CASRN 115-
96-8) in F344/N Rats and B6C3F1 Mice (Gavage
Studies) (1991)
International
Organisation for Economic Co-operation and
Development (OECD), Cooperative Chemicals
Assessment Program (CoCAP)
SIDS initial assessment profile for SIAM 23: Tris(2-
chloroethvDphosphate (CAS no. 115-96-8) (2006)
International Agency for Research on Cancer (IARC)
Monographs on the Evaluation of Carcinogenic Risks to
Humans Volume 71 (1999)
European Union, European Chemicals Agency (ECHA)
European Union Risk Assessment Report: CAS: 115-
96-8: Tris (2-chloroethvl) phosphate. TCEP (2009)
Government of Canada, Environment Canada, Health
Canada
Screening Assessment for the Challenge Ethanol. 2-
chloro-. phosphate (3:1) (Tris(2-chlrorethvl) phosphate
TTCEP1) (2009)
Page 446 of 638
-------
Authoring Organization
Publication
National Industrial Chemicals Notification and
Assessment Scheme (NICNAS), Australian
Government
Ethanol. 2-chloro-. phosphate (3:1): Human health tier
II assessment (2016). and Ethanol. 2-chloro-. phosphate
(3:1): Human health tier III assessment (2017)
Page 447 of 638
-------
Appendix C LIST OF SUPPLEMENTAL DOCUMENTS
Appendix C incudes a list and citations for all supplemental documents included in the Risk Evaluation
for TCEP. See Docket EPA-HQ-QPPT-2018-0476 for all publicly released files associated with this risk
evaluation package; see Docket EPA-HQ-OPPT-2023-0265 for all publicly released files associated
with peer review and public comments.
Associated Systematic Review Protocol and Data Quality Evaluation and Data Extraction
Documents - Provide additional detail and information on systematic review methodologies used as
well as the data quality evaluations and extractions criteria and results.
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review Protocol (U.S. EPA.
2024p) - In lieu of an update to the Draft Systematic Review Protocol Supporting TSCA Risk
Evaluations for Chemical Substances, also referred to as the "2021 Draft Systematic Review
Protocol" (U.S. EPA 202la), this systematic review protocol for the Risk Evaluation for TCEP
describes some clarifications and different approaches that were implemented than those described
in the 2021 Draft Systematic Review Protocol in response to (1) SACC comments, (2) public
comments, or (3) to reflect chemical-specific risk evaluation needs. This supplemental file may also
be referred to as the "TCEP Systematic Review Protocol."
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review Supplemental File:
Data Quality Evaluation and Data Extraction Information for Physical and Chemical Properties
(U.S. EPA. 2024v) - Provides a compilation of tables for the data extraction and data quality
evaluation information for TCEP. Each table shows the data point, set, or information element that
was extracted and evaluated from a data source that has information relevant for the evaluation of
physical and chemical properties. This supplemental file may also be referred to as the "TCEP Data
Quality Evaluation and Data Extraction Information for Physical and Chemical Properties."
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review Supplemental File:
Data Quality Evaluation and Data Extraction Information for Environmental Fate and Transport
(U.S. EPA. 2024t) - Provides a compilation of tables for the data extraction and data quality
evaluation information for TCEP. Each table shows the data point, set, or information element that
was extracted and evaluated from a data source that has information relevant for the evaluation for
Environmental Fate and Transport. This supplemental file may also be referred to as the "TCEP Data
Quality Evaluation and Data Extraction Information for Environmental Fate and Transport."
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review Supplemental File:
Data Quality Evaluation and Data Extraction Information for Environmental Release and
Occupational Exposure (U.S. EPA. 2024u) - Provides a compilation of tables for the data extraction
and data quality evaluation information for TCEP. Each table shows the data point, set, or
information element that was extracted and evaluated from a data source that has information
relevant for the evaluation of environmental release and occupational exposure. This supplemental
file may also be referred to as the "TCEP Data Quality Evaluation and Data Extraction Information
for Environmental Release and Occupational Exposure."
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review Supplemental File:
Data Quality Evaluation and Data Extraction Information for Dermal Absorption (U.S. EPA.
2024s) - Provides a compilation of tables for the data extraction and data quality evaluation
information for TCEP. Each table shows the data point, set, or information element that was
extracted and evaluated from a data source that has information relevant for the evaluation for
Page 448 of 638
-------
Dermal Absorption. This supplemental file may also be referred to as the "TCEP Data Quality
Evaluation and Data Extraction Information for Dermal Absorption."
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review Supplemental File:
Data Quality Evaluation Information for General Population, Consumer, and Environmental
Exposure. (U.S. EPA 2024x) - Provides a compilation of tables for the data quality evaluation
information for TCEP. Each table shows the data point, set, or information element that was
evaluated from a data source that has information relevant for the evaluation of general population,
consumer, and environmental exposure. This supplemental file may also be referred to as the "TCEP
Data Quality Evaluation Information for General Population, Consumer, and Environmental
Exposure."
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review Supplemental File:
Data Extraction Information for General Population, Consumer, and Environmental Exposure (U.S.
EPA 2024r) - Provides a compilation of tables for the data extraction for TCEP. Each table shows
the data point, set, or information element that was extracted from a data source that has information
relevant for the evaluation of general population, consumer, and environmental exposure. This
supplemental file may also be referred to as the "TCEP Data Extraction Information for General
Population, Consumer, and Environmental Exposure."
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review Supplemental File:
Data Quality Evaluation Information for Human Health Hazard Epidemiology (U.S. EPA. 2024z) -
Provides a compilation of tables for the data quality evaluation information for TCEP. Each table
shows the data point, set, or information element that was evaluated from a data source that has
information relevant for the evaluation of epidemiological information. This supplemental file may
also be referred to as the "TCEP Data Quality Evaluation Information for Human Health Hazard
Epidemiology."
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review Supplemental File:
Data Quality Evaluation Information for Human Health Hazard Animal Toxicology (U.S. EPA.
2024y) - Provides a compilation of tables for the data quality evaluation information for TCEP.
Each table shows the data point, set, or information element that was evaluated from a data source
that has information relevant for the evaluation of human health hazard animal toxicity information.
This supplemental file may also be referred to as the "TCEP Data Quality Evaluation Information
for Human Health Hazard Animal Toxicology."
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review Supplemental File:
Data Quality Evaluation Information for Environmental Hazard (U.S. EPA. 2024w) - Provides a
compilation of tables for the data quality evaluation information for TCEP. Each table shows the
data point, set, or information element that was evaluated from a data source that has information
relevant for the evaluation of environmental hazard toxicity information. This supplemental file may
also be referred to as the "TCEP Data Quality Evaluation Information for Environmental Hazard."
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Systematic Review Supplemental File:
Data Extraction Information for Environmental Hazard and Human Health Hazard Animal
Toxicology and Epidemiology (U.S. EPA. 2024q) - Provides a compilation of tables for the data
extraction for TCEP. Each table shows the data point, set, or information element that was extracted
from a data source that has information relevant for the evaluation of environmental hazard and
human health hazard animal toxicology and epidemiology information. This supplemental file may
Page 449 of 638
-------
also be referred to as the "TCEP Data Extraction Information for Environmental Hazard and Human
Health Hazard Animal Toxicology and Epidemiology."
Associated Supplemental Information Documents - Provide additional details and information on
exposure, hazard, and risk assessments.
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File:
Supplemental Information on Environmental Release and Occupational Exposure Assessment
(U.S. EPA 2024n).
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File: E-
FAST Modeling Results (U.S. EPA. 2024a).
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File:
IIOAC Modeling Input and Results (U.S. EPA. 20241).
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File:
Environmental Monitoring Concentrations Reported by Media Type (U.S. EPA. 20240.
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File:
Environmental Monitoring and Biomonitoring Concentrations Summary Table (U.S. EPA.
2024h).
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File:
Consumer Exposure Modeling Inputs (U.S. EPA. 2024e).
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental File Folder:
Supplemental Information on Consumer Exposure Modeling Results (U.S. EPA. 2024f).
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File:
Human Health Hazard Points of Departure Comparison Tables (U.S. EPA. 2024k) - Provides
an Excel spreadsheet of PODs for all studies and hazard outcomes resulting in likely or
suggestive evidence integration conclusions. Basic study details as well as the PODs from each
study and associated HEDs, HECs, and total UFs for non-cancer endpoints, as well as CSFs and
IURs for cancer endpoints are presented.
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File:
Benchmark Dose Modeling Results for TCEP (U.S. EPA. 2024c) - Provides inputs to BMD
modeling as well as outputs for individual health effects associated with hazard outcomes that
have likely evidence integration conclusions. Information includes goodness of fit details for all
models that were run, as well as BMD and BMDL values for the selected BMR and any
comparison BMRs. Graphs of the chosen models are also presented.
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File:
Risk Calculator for Occupational Exposures (U.S. EPA. 2024m).
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File:
Exposure Air Concentration Risk Calculations (U.S. EPA. 2024i).
Page 450 of 638
-------
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File:
Water Quality Portal Processed Water Data (U.S. EPA. 2024o).
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental File Folder:
Supplemental Information on Human Milk PBPK Verner Modeling Results (U.S. EPA. 2024a)
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental File Folder: Biosolids
Screening Tool Modeling Results (U.S. EPA. 2024cT)
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental File Folder:
Consumer Modeling Risk Calculations, Sensitivity Analysis, and Visualizations of Results and
Environmental Water Quality Portal Data (U.S. EPA. 2024b)
Page 451 of 638
-------
Appendix D CONDITIONS OF USE DESCRIPTIONS
The following descriptions are intended to include examples of uses so as not to exclude other activities
that may also be included in the COUs of the chemical substance. To better describe the TSCA COU,
EPA considered CDR submissions from the last two CDR cycles for DINP (CASRN 28553-12-0 and
CASRN 68515-48-0), and the COU descriptions reflect what EPA identified as the best fit for that
submission.
D.l Manufacturing (Import)
Import refers to the import of TCEP into the customs territory of the United States. This condition of use
includes loading/unloading and repackaging (but not transport) associated with the import of TCEP. In
general, chemicals may be imported into the United States in bulk via water, air, land, and intermodal
shipments. These shipments take the form of oceangoing chemical tankers, rail cars, tank trucks, and
intermodal tank containers (U.S. EPA 2021c).
Examples of CDR Submissions. In 2016, one company, reported importation of TCEP (CASRN 115-
96-8) in wet-solid form with a production volume of 158,728 lb. However, the company ceased
operations in 2019 (Aceto US LLC. 2024). There were no reporters of TCEP in the 2020 CDR.
D.2 Processing - Incorporation into Formulation, Mixture, or Reaction
Product - Paint and Coating Manufacturing
This COU refers to the preparation of a product; that is, the incorporation of TCEP into formulation,
mixture, or a reaction product which occurs when a chemical substance is added to a product (or product
mixture), after its manufacture, for distribution in commerce, in this case—processing TCEP into
coating products for commercial (non-consumer) use, including waterborne coatings and resin-/solvent-
based coatings. The general processes for the formulation of waterborne coatings and resin/solvent -
based coatings is similar but with shorter blending/mixing times for the resin-/solvent-based products.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR. However, EPA has met with stakeholders and received comment indicating continued
use of TCEP in paints and coatings used in the aerospace and automotive industries (Alliance for
Automotive Innovation. 2023a. b; FCC. 2021; AIA. 2019).
D.3 Processing - Incorporation into Formulation, Mixture, or Reaction
Product - Polymers used in Aerospace Equipment and Products
This COU refers to the preparation of a product; that is, the incorporation of TCEP into formulation,
mixture, or a reaction product which occurs when a chemical substance is added to a product (or product
mixture), after its manufacture, for distribution in commerce, in this case—processing TCEP into
formulations of aerospace products. In aerospace products, TCEP is used as a flame-retardant additive
component of 2-part polymer and prepolymer resin systems used in potting and casting applications; as
an additive plasticizer and viscosity regulatory with flame-retarding properties for polyurethane,
polyesters, polyvinyl chloride, and other polymers; in the production of unsaturated polyester resins and
in acrylic resins and coatings; and for production of polyurethane foam.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR. However, stakeholders provided information regarding the use of TCEP in polymers
used in the aerospace industry (AIA. 2021a. b, 2020. 2019).
Page 452 of 638
-------
D.4 Processing - Incorporation into Article - Aerospace Equipment and
Products and Automotive Articles and Replacement Parts Containing
TCEP
This COU refers to the preparation of an article; that is, the incorporation of TCEP into articles, meaning
TCEP becomes a component of the article, after its manufacture, for distribution in commerce. In this
case, TCEP is present as a flame-retardant and plasticizer additive in polymer resins used in potting and
casting applications in the aerospace industry as well as for production of polyurethane foam in aircraft
and aerospace products. EPA identified that plastic products with TCEP-containing cured paints and
coatings are currently used (via incorporation) in articles for automotive applications and that the TCEP-
containing foam products are currently used (via incorporation) in articles for aircraft and aerospace
applications. Specific aerospace industrial uses include resins and elastomeric coatings, polyurethane
casting for aircraft interiors and as a flame retardant for aircraft furniture. Specific automotive
processing uses include installation of products and replacement parts with cured TCEP-containing
paints and coatings into automobiles.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR. However, EPA has received public comment that TCEP may be processed into
automotive articles and replacement parts (Alliance for Automotive Innovation. 2023b).
D.5 Processing - Recycling
This COU refers to the process of treating generated waste streams (i.e., which would otherwise be
disposed of as waste), containing TCEP, which are collected, either on-site or transported to a third-
party site, for commercial purpose. TCEP may be present as an additive in components of electronics
and electrical equipment that is recycled. E-waste recycling activities include receiving e-waste at the
facility, dismantling or shredding the e-waste, and sorting the recycled articles and generated scrap
materials. TCEP has been identified in articles that are commonly recycled such as insulation, plastics,
and foam. TCEP may be present within flexible foam, fabric, textile, and other applications that have
been made from recycled foam scraps generated during trimming of original TCEP-containing
manufactured foam products.
Examples of CDR Submissions. EPA notes that although TCEP was not reported for recycling in the
2016 or 2020 CDR reporting periods, the Agency is assuming that recycling waste streams likely
contain TCEP.
D.6 Distribution in Commerce
For purposes of assessment in this risk evaluation, distribution in commerce consists of the
transportation associated with the moving of TCEP or TCEP-containing products between sites
manufacturing, processing, or recycling TCEP or TCEP-containing products, or to final use sites, or for
final disposal of TCEP or TCEP-containing products. More broadly under TSCA, "distribution in
commerce" and "distribute in commerce" are defined under TSCA section 3(5).
D.7 Industrial Use - Other Use - Aerospace Equipment and Products and
Automotive Articles and Replacement Parts Containing TCEP
This COU is referring to the industrial use of products or articles containing TCEP in aerospace
equipment and automotive articles and replacement parts. Meaning the use of TCEP after it has already
been incorporated into a product or article, as opposed to when it is used upstream (e.g., when TCEP is
processed into the product or article). TCEP is present in the cured resin or foam components of articles
Page 453 of 638
-------
that are installed in aircrafts or aerospace vehicles and in the cured paints and coatings in automotive
articles and replacement parts. Examples of possible TCEP uses in aircraft and aerospace products
include its presence as a flame retardant in aircraft furniture foams, electronics, or structural
components. TCEP-containing articles are used as received at the site, with minimal or no reshaping or
processing of the article prior to manual installation into the aircraft or aerospace vehicle and manual
and/or robotic installation into the automobile. Industrial use of TCEP for this COU involves
maintenance of aerospace equipment and products and workers handling automotive articles and
replacement parts at automotive manufacturers.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR. However, EPA has received comment that TCEP is used in the aerospace industry as
well as in automotive articles and replacement parts (Alliance for Automotive Innovation. 2023b;
Boeing. 2023; AIA. 2021a. 2019V
D.8 Industrial Use - Paints and Coatings
This COU is referring to the use of TCEP in various industrial sectors as a component of industrial
paints and coatings. Meaning the use of TCEP after it has already been incorporated into a paint or
coating product or mixture, as opposed to when it is used upstream (e.g., when TCEP is processed into
the paint or coating formulation). TCEP is an additive component in paints and coatings for industrial
and commercial use as a flame-retardant coating to achieve flame spread or fire protection standards for
automobile articles and replacement parts. Products include TCEP-containing paints and coatings that
are applied in an industrial setting to automobile bodies. Coating application methods for TCEP-
containing paints and coatings in the automotive industry include robotic and manual applications.
TCEP will remain in the coating as an additive in the dried/cured coating on the substrate.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR. However, EPA has received comment indicating continued use of TCEP in paints and
coatings used in the aerospace and automotive industries (Alliance for Automotive Innovation. 2023a. b;
AIA. 2019V
D.9 Commercial Use - Other Use - Aerospace Equipment and Products
and Automotive Articles and Replacement Parts Containing TCEP
This COU is referring to the commercial use of TCEP in aerospace equipment and products and
automotive articles and replacement parts, which already have TCEP incorporated into them. Meaning
the use of TCEP-containing aerospace equipment and products and automotive articles and replacement
parts in a commercial setting, such as a worker operating in unoccupied parts of airplanes or driving a
vehicle or an automotive parts business, as opposed to upstream use of TCEP (e.g., when TCEP
containing products are used in the manufacturing of the airplane or automotive) or use in an industrial
setting. TCEP is present in the cured resin in aerospace equipment and products, the cured paints and
coatings in automotive articles and replacement parts and in foam components of articles that are
installed in aircrafts or aerospace vehicles. Examples of possible TCEP uses in aircraft and aerospace
products include its presence as a flame retardant in aircraft furniture foams, electronics or structural
components. TCEP-containing articles are used as received at the site, with minimal or no reshaping or
processing of the article prior to manual installation into the aircraft or aerospace vehicle and manual
installation into the automobile. Commercial use of TCEP for this COU involves use of aerospace
equipment or products by airliners and workers (e.g. auto repair shops and commercial drivers) handling
automotive articles and replacement parts.
Page 454 of 638
-------
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR. However, EPA has received comment that TCEP is used in the aerospace industry as
well as in automotive articles and replacement parts (Alliance for Automotive Innovation. 2023b;
Boeing. 2023; AIA. 2021a. 2019V
D.10 Commercial Use - Paints and Coatings
This COU is referring to the commercial use of TCEP already incorporated as a flame retardant in paints
and coatings. TCEP is an additive component in paints and coatings for industrial and commercial use as
a flame-retardant coating to achieve flame spread or fire protection standards for structural and electrical
components and include waterborne coatings and resin-/solvent-based coatings. Products include 1-part
coatings and 2-part epoxy resins that are typically used on electrical cables, exterior (masonry) surfaces,
or unoccupied parts of a building such as mechanical rooms, attics, and crawl spaces that may contain
foam that needs to be coated in flame retardants. Other applications include coating large industrial steel
or aluminum structures. Coating application methods for TCEP-containing paints and coatings include
spray gun, brush, and trowel coating for use on structures or equipment. TCEP will remain in the coating
as an additive in the dried/cured coating on the substrate. These products would be purchased by
commercial operations and applied by professional contractors in various commercial settings.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR. However, EPA has received comment indicating use of TCEP in paints and coatings
used in the aerospace and automotive industries (Alliance for Automotive Innovation. 2023a. b; AIA.
2019).
D.ll Commercial Use - Laboratory Chemicals
This COU is referring to the commercial use of TCEP in laboratory chemicals. TCEP is used as a
laboratory chemical, such as in a chemical standard or reference material during analyses. The users of
products under this category would be expected to apply these products through general laboratory use
applications. Commercial use of laboratory chemicals may involve handling TCEP by hand-pouring or
pipette and either adding to the appropriate labware in its pure form to be diluted later or added to dilute
other chemicals already in the labware. Laboratory TCEP products are pure TCEP in neat liquid form.
The Agency notes that the same applications and methods used for quality control can be applied in
industrial and commercial settings.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR. However, EPA has received public comment indicating laboratory use of TCEP (NASA.
2020).
D.ll Commercial Use - Furnishing, Cleaning, Treatment/Care Products
- Fabric and Textile Products
This COU is referring to the commercial use of TCEP already incorporated as a flame-retardant
plasticizer in fabric and textile products. TCEP was previously used as a flame-retardant plasticizer in
unsaturated polyester resins, which was used in polyester yarn and fabric products including clothing,
bed sheets, blankets, upholstered furniture, industrial polyester fibers, fabrics for conveyor and safety
belts, and coating fabric. EPA has not found evidence of TCEP currently being used in fabric and textile
products for both commercial and consumer uses. The Agency has confirmed from literature sources
that TCEP was used for these purposes in the past but was phased out of these uses starting in the late
1980s or early 1990s in favor of other flame retardants or flame-retardant formulations. Because
Page 455 of 638
-------
manufacturing and processing of TCEP is not ongoing, this COU focuses on the end of service life
disposal for human health and the environment.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR.
D.13 Commercial Use - Furnishing, Cleaning, Treatment/Care Products
- Foam Seating and Bedding Products
This COU is referring to the commercial use of TCEP already incorporated in foam seating and bedding
products and furnishings. EPA has confirmed that the manufacturing and processing of TCEP into foam
seating and bedding products for commercial and industrial use, outside of aircrafts and aerospace
products, has been phased out. TCEP was previously used in flexible urethane cushions that were used
in institutional mattresses, furniture foam padding, automotive seat cushions and padding, carpet
underlay, and pillow and mattress padding. EPA identified during scoping that foams for furniture was a
major use of TCEP prior to the 1990s. The Agency has confirmed from literature sources that TCEP was
used for these purposes in the past but was phased out of these uses starting in the late 1980s or early
1990s in favor of other flame retardants or flame-retardant formulations. Because manufacturing and
processing of TCEP is not ongoing, this COU focuses on the end of service life disposal for human
health and the environment.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR.
D.14 Commercial Use - Construction, Paint, Electrical, and Metal
Products - Building/Construction Materials - Insulation
This COU is referring to the commercial use of TCEP in commercial sectors associated with
construction products that contain TCEP as a plasticizer or flame retardant, such as at a business or at a
job site as opposed to upstream use of TCEP (e.g., when TCEP is processed into the construction
material) or use in an industrial setting. Rigid polyurethane foams for insulation, specifically
commercial roofing insulation, was identified as a potential application for TCEP and was identified as a
major use of TCEP prior to the 1990s. EPA has not found any modern evidence that TCEP is still being
manufactured or processed for incorporation into building and construction materials. The Agency has
confirmed from literature sources that TCEP was used for these purposes in the past but was phased out
of these uses starting in the late 1980s or early 1990s in favor of other flame retardants or flame-
retardant formulations. Because manufacturing and processing of TCEP is not ongoing, this COU
focuses on the end of service life disposal for human health and the environment.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR.
D.15 Commercial Use - Construction, Paint, Electrical, and Metal
Products - Building/Construction Materials - Wood and Engineered
Wood Products - Wood Resin Composites
This COU is referring to the commercial use of TCEP in commercial sectors associated with
construction products that contain TCEP as a plasticizer or flame retardant, such as at a business or at a
job site as opposed to upstream use of TCEP (e.g., when TCEP is processed into the construction
material) or use in an industrial setting. TCEP was previously incorporated into construction, paint,
Page 456 of 638
-------
electrical, and metal products, such as roofing insulation and rigid foam. It is possible that TCEP was
used in the resins that bond wood products together. TCEP use in engineered wood products was a
minor use in only niche products, such as furniture production as opposed to larger scale uses in building
construction. EPA has not found any modern evidence that TCEP is still being manufactured or
processed for incorporation into building and construction materials. The Agency has confirmed from
literature sources that TCEP was used for these purposes in the past but was phased out of these uses
starting in the late 1980s or early 1990s in favor of other flame retardants or flame-retardant
formulations. Because manufacturing and processing of TCEP is not ongoing, this COU focuses on the
end of service life disposal for human health and the environment.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR.
D.16 Consumer Use - Paints and Coatings, Including Those Found on
Automotive Articles and Replacement Parts
This COU is referring to the consumer use of TCEP already incorporated as a flame retardant in paints
and coatings. TCEP-containing paints and coatings are unlikely to be available for purchase by
consumers, and EPA did not find any evidence indicating the continued use of TCEP-containing paints
and coatings by consumers. However, the Agency has found that some automotive articles and
replacement parts painted with TCEP-containing paint may be available for purchase and installation by
DIY users. Consumers may be exposed to TCEP if entering unoccupied parts of buildings such as attics
or crawl spaces that might contain electrical cables or foams treated paints and coatings that may contain
TCEP. Consumers can also be exposed to TCEP if using automobiles containing articles or replacement
parts painted with TCEP-containing paint.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR. However, EPA has met with stakeholders who report the use of TCEP-containing paints
and coatings on automotive articles and replacement parts that may be accessible to consumers (Alliance
for Automotive Innovation. 2023a. b).
D.17 Consumer Use - Furnishing, Cleaning, Treatment/Care Products -
Fabric and Textile Products
This COU is referring to the consumer use of TCEP already incorporated as a flame-retardant plasticizer
in fabric and textile products. TCEP was previously used as a flame-retardant plasticizer in unsaturated
polyester resins, which was used in polyester yarn and fabric products including clothing, bed sheets,
blankets, upholstered furniture, industrial polyester fibers, fabrics for conveyor and safety belts, and
coating fabric for carpets and upholstery. EPA has not found modern evidence of TCEP currently being
used in fabric and textile products for both commercial and consumer uses.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR.
D.18 Consumer Use - Furnishing, Cleaning, Treatment/Care Products -
Foam Seating and Bedding Products
This COU is referring to the consumer use of TCEP already incorporated in foam seating and bedding
products and furnishings. EPA has confirmed that the manufacturing and processing of TCEP into foam
seating and bedding products for commercial and industrial use, outside of aircrafts and aerospace
products, has been phased out. TCEP was previously used in flexible urethane cushions that were used
Page 457 of 638
-------
in institutional mattresses, furniture foam padding, automotive seat cushions and padding, carpet
underlay, and pillow and mattress padding.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR.
D.19 Consumer Use - Construction, Paint, Electrical, and Metal
Products - Building/Construction Materials - Insulation
This COU is referring to the consumer use of TCEP associated with construction products that contain
TCEP as a plasticizer or flame retardant. TCEP was previously incorporated into construction, paint,
electrical, and metal products, such as roofing insulation and rigid foam. EPA has not found any modern
evidence that TCEP is still being manufactured or processed for incorporation into building and
construction materials.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR.
D.20 Consumer Use - Construction, Paint, Electrical, and Metal
Products - Building/Construction Materials - Wood and Engineered
Wood Products - Wood Resin Composites
This COU is referring to the consumer use of TCEP associated with construction products that contain
TCEP as a plasticizer or flame retardant. TCEP was previously incorporated into construction, paint,
electrical, and metal products, such as roofing insulation and rigid foam. It is possible that TCEP was
used in the resins that bond wood products together. TCEP use in engineered wood products was a
minor use in only niche products, such as furniture production as opposed to larger scale uses in building
construction. EPA has not found any modern evidence that TCEP is still being manufactured or
processed for incorporation into building and construction materials.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR.
D.21 Disposal
Each of the COUs of TCEP may generate waste streams of the chemical. For purposes of the TCEP risk
evaluation, this COU refers to the TCEP in a waste stream that is collected and transported to third-party
sites for disposal or treatment. This COU also encompasses TCEP contained in wastewater discharged
to POTW or other, non-public treatment works for treatment, and other wastes. TCEP is expected to be
released to other environmental media, such as introductions of biosolids to soil or migration to water
sources, through waste disposal (e.g., disposal of formulations containing TCEP, plastic and rubber
products, textiles, and transport containers). Disposal may also include destruction and removal by
incineration (U.S. EPA 2021b). Recycling of TCEP and TCEP containing products is considered a
different COU. Environmental releases from industrial sites are assessed in each COU. Disposal of
TCEP occurs during the disposal of TCEP-containing articles, such as furniture, clothing, and plastic
waste, as well as during demolition of and disposal of building materials that may contain TCEP, such
as roofing insulation.
Examples of CDR Submissions. There were no reporters for TCEP for this use in neither the 2016 nor
the 2020 CDR.
Page 458 of 638
-------
Appendix E DETAILED EVALUATION OF POTENTIALLY
EXPOSED OR SUSCEPTIBLE SUBPOPULATIONS
E.l PESS Based on Greater Exposure
In this section, EPA addresses the following potentially exposed populations expected to have greater
exposure to TCEP. Table Apx E-l presents the quantitative data sources that were used in the PESS
exposure analysis for incorporating increased background and COU-specific exposures.
Page 459 of 638
-------
Table Apx E-l. PESS Evidence Crosswalk for Increased Exposure
Category
Subcategory
Increased Background Exposure
Increased COU or Pathway Specific
Exposures
Quantitative Data Sources
Lifestage
Embryo/fetus
• Transfer of exposure from the parent
(placenta to fetus)
• Ratio of placenta: maternal serum (Rpm)
concentrations shown to range from 0.76
for TCEP
• (Wane et al.. 2021)
Children
(infants, toddlers)
• EPA did not identify sources of increased
background exposure anticipated for this
lifestage
• Hand to mouth behavior leads to
increased ingestion of household dust
• Age-appropriate behavior patterns
(elevated soil ingestion exposure
(children's activities with soil, children
playing mud)
• Human milk exposure from maternal
doses derived from TSCA sources
• Different exposure factors
• Drinking water exposure from TSCA
sources
• EPA Age Grouping
Guidance
• Exposure Factors
Handbook (U.S. EPA.
2017d)
• See Section 5.1.3.4.7
Geriatric
• Older populations that generally use
supplements may be at higher exposure to
TCEP due to use of Fish oil supplements
• EPA did not identify sources of
increased COU or pathway specific
exposure for this lifestage
• (Poma et al.. 2018)
Sociodemo-
graphic/
Lifestyle
Race/Ethnicity
• EPA did not identify sources of increased
background exposure anticipated for this
lifestage
• TCEP levels in dust are significantly
associated with the presence of
extremely worn carpets; lower
socioeconomic status (SES) populations
are more prone to having homes with
older carpets due to their cost of
replacement
• Fenceline populations (typically lower
SES) may live closer to emitting sources
• (Castorina et al.. 2017)
Subsistence
Fishing
• EPA did not identify sources of increased
background exposure anticipated for this
lifestage
• Subsistence fishing populations that
consumer more fish have elevated levels
of TCEP exposure
• See Section 5.1.3.4.3
Occupational
Firefighters
• Firefighters may be at increased risk of
TCEP exposures during structure fires
(Maver et al.. 2021).
• EPA did not identify sources of
increased COU or pathway specific
exposure for firefighters
• See qualitative discussion
Section 5.3.3
• (Javatilaka et al.. 2017).
Page 460 of 638
-------
Category
Subcategory
Increased Background Exposure
Increased COU or Pathway Specific
Exposures
Quantitative Data Sources
Consumer
High frequency
consumers
• Non-TSCA source such as dietary
exposures through food, food packaging,
drugs, and personal care products that
contain TCEP
• Consumer products designed for
children (e.g., children's outdoor play
structures, toy foam blocks) may lead to
elevated exposures for children and
infants.
• Use Report
• EPA's Exposure Factors
Handbook (Ch. 17)
• See Sections 5.1.2.2 and
5.1.3.4.8
High duration
consumers
Page 461 of 638
-------
E.2 PESS Based on Greater Susceptibility
In this section, EPA addresses subpopulations expected to be more susceptible to TCEP exposure than
other populations. TableApx E-2 presents the data sources that were used in the PESS analysis
evaluating susceptible subpopulations and identifies whether and how the subpopulation was addressed
quantitatively in the risk evaluation of TCEP.
Some observations may be made regarding factors that may increase susceptibility to the effects of
TCEP. Human data are available on health effects of TCEP that may suggest there are susceptible
subpopulations, although as identified in Section 5.2.3, human evidence is only slight (with evidence as
indeterminate for thyroid effects). Percy et al. (2022) found increased BCEP in urine was associated
with lower IQ in children with low socioeconomic status (SES) using more than one measure related to
SES. In contrast, for the full sample of children (a wider range of economic status), the authors found a
small positive (but not statistically significant) relationship between BCEP IQ (Percy et al.. 2022).
Similarly, in a group of children across SES levels, maternal BCEP exposure was marginally associated
with higher IQ (Percy et al.. 2021).
Human epidemiological studies show slight evidence for developmental effects related to growth and
gestational age examining associations with a TCEP metabolite (BCEP) in mother's urine during
gestation. Both sexes and male offspring alone who were less likely to be small for their gestational age
in one study (Oh et al.. 2024). Increased BCEP was also associated with increased skinfold thickness in
male offspring (two measures of thickness) and both sexes (one measure) (Crawford et al.. 2020).
Female children had a greater incidence of being pre-term (Oh et al.. 2024) and lower birthweight and
length (Yang et al.. 2022).
Effects may differ by gender, as identified by some epidemiological studies. As noted above, female
developmental outcomes related to growth and gestational age may differ from males. Mendv et al.
(2024) found that TCEP in dust had a more pronounced effect on hay fever and allergies in female
children compared with males. Also, TCEP in serum was associated with more statistically significant
changes in thyroid hormones in females than in males (Liu et al.. 2022) but evidence is indeterminate.
Animal studies identified developmental effects (NTP. 1991a) as well as sensitive sexes for certain
health outcomes—higher incidence of neurotoxicity in female rats (NTP. 1991b) and greater sensitivity
of male (vs. female) mice in reproductive effects (Chen et al.. 2015a)—and EPA quantified risks based
on these endpoints in the risk evaluation. It is possible that these differences in rodents reflect
differences in humans. However, if sex differences in susceptibility among rodents are due solely to
differences in toxicokinetics, there is uncertainty for humans given a lack of metabolic differences
among sexes in experiments using human liver tissues (Chapman et al.. 1991).
As identified in Table Apx E-2, many other susceptibility factors that are generally considered to
increase susceptibility of individuals to chemical hazards. These factors include pre-existing diseases,
alcohol use, diet, stress, among others. The effect of these factors on susceptibility to health effects of
TCEP is not known; therefore, EPA is uncertain about the magnitude of any possible increased risk from
effects associated with TCEP exposure.
For non-cancer endpoints, EPA used a default value of 10 for human variability (UFh) to account for
increased susceptibility when quantifying risks from exposure to TCEP. The Risk Assessment Forum, in
A Review of the Reference Dose and Reference Concentration Processes (U.S. EPA. 2002b). discusses
some of the evidence for choosing the default factor of 10 when data are lacking and describe the types
Page 462 of 638
-------
of populations that may be more susceptible, including different lifestages (e.g., of children and elderly).
U.S. EPA (2002b). however, did not discuss all the factors presented in Table Apx E-2. Thus,
uncertainty remains regarding whether these additional susceptibility factors would be covered by the
default UFh value of 10 chosen for use in the TCEP risk evaluation. In addition, given that EPA is using
a default UFh in the absence of data regarding whether adverse effects from TCEP exposure differ for
certain subpopulations (such as those with genetic polymorphisms or underlying diseases), it is also not
known whether the chosen default UFh would fully cover pre-existing diseases or disorders U.S. EPA
(2002bY
For cancer, the dose-response model applied to animal tumor data employed low-dose linear
extrapolation, and this assumes any TCEP exposure is associated with some positive risk of getting
cancer. EPA made this assumption in the absence of an established MOA for TCEP and according to
guidance from U.S. EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005b). Assuming
all TCEP exposure is associated with some risk is likely to be health conservative because EPA does not
believe that a mutagenic MOA is likely for TCEP and a threshold below which cancer does not occur is
expected to exist. However, information is lacking with which to determine an appropriate threshold.
Even though the cancer dose-response modeling assumes any exposure is associated with a certain risk,
EPA presents risk estimates in comparison with benchmark risk levels (1 in 1,000,000 to 1 in 10,000).
Although there is likely to be variability in susceptibility across the human population, EPA did not
identify specific human groups that are expected to be more susceptible to cancer following TCEP
exposure. Other than relying on animal tumor data for the more sensitive sex, the available evidence
does not allow EPA to evaluate or quantify the potential for increased cancer risk in specific
subpopulations, such as for individuals with pre-existing diseases or those who smoke cigarettes. Given
that a mutagenic mode of action is unlikely, EPA does not anticipate greater cancer risks from early life
exposure to TCEP. Therefore, EPA is not applying an age-dependent adjustment factor.
Page 463 of 638
-------
Table Apx
1-2. PESS Evidence Crosswalk for Biological Suscepi
tibility Considerations
Susceptibility
Category
Examples of
Specific
Factors
Direct Evidence this Factor
Modifies Susceptibility to TCEP
Indirect Evidence of Interaction with Target Organs or
Biological Pathways Relevant to TCEP
Susceptibility
Addressed in Risk
Evaluation?
Description of Interaction
Key Citations
Description of Interaction
Key Citation(s)
Lifestage
Embryos/
fetuses/infants
Slight evidence for
growth/gestational age effects
in humans.
Slight animal evidence for
developmental toxicity (e.g.,
decreased fertility and live
births with some increased
severity in the second
generation)
Lack of uniquely sensitive
effects on neurodevelopment
in animals (doses up to 90
mg/kg-day)
Oh et al. (2024)
Crawford et al.
(2020)
Yang et al. (2022)
NTP (1991a)
Moser et al. (2015)
POD for male
reproductive endpoints
protective of effects in
offspring "
Pregnancy/
lactating status
Rodent dams not particularly
susceptible during pregnancy
and lactation except in one
prenatal study, in which 7 of
30 dams died at 200 mg/kg-
day
NTP (1991a)
Hazleton
Laboratories (1983)
Moser et al. (2015)
Kawashima et al.
(1983)
POD for male
reproductive endpoints
protective of effects in
dams
Males of
reproductive
age
Reproductive outcomes
(effects on seminiferous
tubules) in adolescent male
mice
Chen et al. (2015a)
Possible contributors to male
reproductive effects/infertility (see
also factors in other rows):
• Enlarged veins of testes
• Trauma to testes
• Anabolic steroid or illicit drug
use
• Cancer treatment
CDC (2023b)
POD for this endpoint
and study used to
calculate non-cancer risks
Children
Slight evidence for lower IQ
in children with low SES
Reproductive outcomes
(effects on seminiferous
tubules) in adolescent male
mice
Percv et al. (2022)
Chen et al. (2015a)
Adolescent animal POD
used to calculate non-
cancer risks; other
variability and
uncertainty addressed
using default UFh
Elderly
No direct evidence identified
Use of default UFh
Page 464 of 638
-------
Susceptibility
Category
Examples of
Specific
Factors
Direct Evidence this Factor
Modifies Susceptibility to TCEP
Indirect Evidence of Interaction with Target Organs or
Biological Pathways Relevant to TCEP
Susceptibility
Addressed in Risk
Evaluation?
Description of Interaction
Key Citations
Description of Interaction
Key Citation(s)
Pre-existing
disease or
disorder
Health
outcome/
target organs
No direct evidence identified
Several conditions may contribute
to male reproductive
effects/infertility:
• Hormone disorders
(hypothalamus/ pituitary
glands)
• Diabetes, cystic fibrosis,
autoimmune disorders, certain
infections
Viruses such as human papilloma
virus can increase susceptibility to
cancer
CDC (2023b)
CDC (2023a)
Use of default UFh
Toxicokinetics
Sex differences in
toxicokinetic parameters
might have resulted in
differences in susceptibility.
Herret al. (1991)
Burka et al. (1991)
Chapman et al.
(1991)
Use of PODs for the
more sensitive sex; Use
of default UFh
Lifestyle
activities
Smoking
No direct evidence identified
Heavy smoking may increase
susceptibility for reproductive
outcomes and cancer.
CDC (2023a)
CDC (2023b)
Qualitative discussion in
this section (D.2) and this
table
Alcohol
consumption
No direct evidence identified
Heavy alcohol use may affect
susceptibility to reproductive
outcomes and cancer.
CDC (2023b)
Qualitative discussion in
this section (D.2) and this
table
Physical
Activity
No direct evidence identified
Insufficient activity may increase
susceptibility to multiple health
outcomes.
Overly strenuous activity may also
increase susceptibility.
CDC (2022)
Qualitative discussion in
this section (D.2) and this
table
Page 465 of 638
-------
Susceptibility
Category
Examples of
Specific
Factors
Direct Evidence this Factor
Modifies Susceptibility to TCEP
Indirect Evidence of Interaction with Target Organs or
Biological Pathways Relevant to TCEP
Susceptibility
Addressed in Risk
Evaluation?
Description of Interaction
Key Citations
Description of Interaction
Key Citation(s)
Sociodemo-
graphic status
Race/ethnicity
No direct evidence identified
(e.g., no information on
polymorphisms in TCEP
metabolic pathways or
diseases associated
race/ethnicity that would lead
to increased susceptibility to
effects of TCEP by any
individual group)
Qualitative discussion in
this section (E).2) and this
table
Socioeconomic
status
Slight evidence for lower IQ
in children with low SES
Percv et al. (2022)
Individuals with lower incomes
may have worse health outcomes
due to social needs that are not
met, environmental concerns, and
barriers to health care access.
ODPHP (2023b)
Qualitative discussion in
this section (E).2) and this
table
Sex/gender
Humans: Slight evidence for
some differences
(developmental,
immunotoxicity);
indeterminate for thyroid
Males (mice): Potentially
more sensitive regarding
reproductive effects
Females (rats): More sensitive
for neurotoxicity
Metabolism experiments using
liver slices and microsomes
show differences in
metabolism by sex for rats,
but not for humans. Thus,
there is uncertainty regarding
whether human females and
males are susceptible
subpopulations.
Oh et al. (2024)
Crawford et al.
(2020)
Yang et al. (2022)
PODs are used in the risk
evaluation for both
endpoints.
Mendv et al. (2024)
NTP (1991a)
NTP (1991b)
Chen et al. (2015a)
Chapman et al.
(1991)
Page 466 of 638
-------
Susceptibility
Category
Examples of
Specific
Direct Evidence this Factor
Modifies Susceptibility to TCEP
Indirect Evidence of Interaction with Target Organs or
Biological Pathways Relevant to TCEP
Susceptibility
Addressed in Risk
Factors
Description of Interaction
Key Citations
Description of Interaction
Key Citation(s)
Evaluation?
Diet
No direct evidence identified
Poor diets can lead to chronic
illnesses such as heart disease,
type 2 diabetes, and obesity.
Obesity can increase susceptibility
to cancer.
CDC (2023a)
CDC (2020)
CDC (2023c)
Qualitative discussion in
this section (D.2) and this
table
Nutrition
Malnutrition
No direct evidence identified
Micronutrient malnutrition can
lead to multiple conditions that
include birth defects, maternal and
infant deaths, preterm birth, low
birth weight, poor fetal growth,
childhood blindness, undeveloped
cognitive ability.
Thus, malnutrition may increase
susceptibility to some/all health
outcomes associated with TCEP.
CDC (2021)
CDC (2023c)
Qualitative discussion in
this Section (D.2) and
this table
Genetics/
epigenetics
Target organs
No direct evidence identified
Genetic disorders, such as
Klinefelter's syndrome, Y-
chromosome microdeletion,
myotonic dystrophy can affect
male reproduction/fertility
CDC (2023b)
Use of default UFh to
assess variability among
humans
Toxicokinetics
No direct evidence identified
Specific enzymes have not been
identified for TCEP's metabolic
pathways. Therefore, potential
polymorphisms are not known.
Use of default UFh to
assess variability among
humans
Page 467 of 638
-------
Susceptibility
Category
Examples of
Specific
Factors
Direct Evidence this Factor
Modifies Susceptibility to TCEP
Indirect Evidence of Interaction with Target Organs or
Biological Pathways Relevant to TCEP
Susceptibility
Addressed in Risk
Evaluation?
Description of Interaction
Key Citations
Description of Interaction
Key Citation(s)
Other chemical
and
nonchemical
stressors
Built
environment
No direct evidence identified
Poor-quality housing is associated
with a variety of negative health
outcomes.
ODPHP (2023a)
Qualitative discussion in
this Section (D.2) and
this table
Social
environment
No direct evidence identified
Social isolation and other social
determinants (e.g., decreased
social capital, stress) can lead to
negative health outcomes.
CDC (2023d)
ODPHP (2023c)
Qualitative discussion in
this Section (D.2) and
this table
Chemical co-
exposures
An in vitro study of liver cells
co-exposed to TCEP and
benzo[a]pyrene activated
pathways associated with cell
proliferation and inflammation
and increased expression of
pro-inflammatory cytokines,
whereas exposure to TCEP
alone did not.
TCEP showed anti-estrogenic
activity (32 percent inhibition)
in vitro using the breast
adenocarcinoma cell line,
MCF-7 after co-exposure with
17B-estradiol.
Zhang et al. (2017b)
Qualitative discussion in
this Section (D.2) and
this table
Krivoshiev et al.
(2016)
"An error in reporting the results in NTP (1991a) precluded using sex ratio; use of this endpoint would have resulted in using a LOAEL of 175 mg/kg-dav with an HED of 23.3
mg/kg-day and a benchmark MOE of 300. This would have resulted in similar but slightly greater risk.
Page 468 of 638
-------
Appendix F PHYSICAL AND CHEMICAL PROPERTIES AND
FATE AND TRANSPORT DETAILS
F.l Physical and Chemical Properties Evidence Integration
The physical and chemical property values selected for use in the risk evaluation for TCEP are given in
Table 2-1. These values were taken from the Final Scope of the Risk Evaluation for Tris(2-chloroethyl)
Phosphate (TCEP) CASRN115-96-8 (U.S. EPA. 2020b). except for physical form, vapor density,
autoflammability, flashpoint, Henry's Law constant, and logarithmic octanokair partition coefficient
(log Koa). In the final scope, no vapor density, log Koa, and autoflammability data were reported and a
flashpoint value from a medium-quality study was provided. After the final scope was published,
additional data were identified in the systematic review process.
The systematic review process identified multiple data with the same quality rating for many physical
and chemical properties discussed in this document. Some of these data were duplicates that were
initially extracted more than once (e.g., when multiple databases cite the same study), but were later
removed during data curation before any further analysis. Much of the remaining data were collected
under standard environmental conditions (i.e., 20 to 25 °C and 760 mm Hg).
When a specific data point is cited for a given physical and chemical property, priority is given to data
from expert-curated, peer-reviewed databases that have been identified as "trusted sources" (U.S. EPA.
2021a). If no data are available from trusted databases, second preference is given to measured data
from studies that implement experimental measurements according to established test guidelines or
which are conducted according to scientific principles with sufficient documentation. Finally, estimated,
or calculated data are only presented in the instance that no measured data is available.
A composite plot comprising box and whisker plots of reported high-, medium-, and low-quality
physical and chemical property data values are shown in FigureApx F-l.
Page 469 of 638
-------
«
-40-
O
o
o
O
CD
^ -50-
0
*
13
ra
03
Q_
~
0
o
E
£ -60-
fl
E
0
h-
1
-70-
<
350-
300-
250-
•200-
150-
100 -
1.42-
E
o
"B) 1-41-
c
0
•g 9.775-
03
0
a:
Melting Point
Boiling Point
Density
9.750-
Vapor Density
0.5-
8000-
O)
X
0.4-
E
E
o
03
| 1.4725-
1.4700-
Refractive Index
Figure Apx F-l. Box and Whisker Plots of Reported Physical and Chemical Property Data Values
F.l.l Physical Form
In the final scope (J.S. EPA. 2020b). physical state and physical properties were 2 of 17 endpoints
provided. As provided in the Final Scope of Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP)
Supplemental File - Data Extraction and Data Evaluation Tables for Physical and Chemical Property
Studies (U.S. EPA. 2020c), only one source was identified and evaluated as a high-quality data for the
physical state endpoint. Ultimately, "liquid" was used in the risk evaluation. For physical properties, two
sources were identified and evaluated as high-quality studies. The reason was not provided, but "clear,
transparent liquid" was preferred and reported over "low odor." For this risk evaluation, both endpoints
were combined and re-named to physical form. After the systematic review process was completed, six
high-quality data were identified and extracted while a medium-quality study was excluded. TCEP is
Page 470 of 638
-------
identified as a clear, transparent liquid with slight odor (DOE. 2016; U.S. EPA. 2015b; ECB. 2009;
Lewis and Hawlev. 2007; Weil. 2001). These descriptions agree with the qualitative description given in
the final scope (U.S. EPA. 2020b).
F.1.2 Vapor Pressure
Systematic review identified 19 high-quality vapor pressure data, including 12 data collected at 20-25
°C. However, five studies reported extrapolated vapor pressures from measured vapor pressures at
higher temperatures and those studies were excluded for this risk evaluation. The remaining seven data
collected under standard environmental conditions cover the range of 3.59x 10"4 to 0.062 mm Hg at 20-
25 °C. The vapor pressure of 0.0613 mm Hg at 25 °C reported by Dobry and Keller (1957) was selected
for this risk evaluation because out of the 7 remaining vapor pressure data, 5 data reported the vapor
pressure from Dobry and Keller (1957).
F.1.3 Vapor Density
A vapor density data was identified through systematic review. It was from a secondary source, NCBI
(2020) and rated it high-quality. Therefore, the vapor density of 9.8 was included in the risk evaluation.
The primary source of the data is ILO (2019).
F.1.4 Water Solubility
Systematic review identified 19 high-quality water solubility data, including 9 data collected at 20 °C.
The data collected under standard environmental conditions cover the range of 7,000-7,900 mg/L at 20
°C. The water solubility value of 7,820 mg/L at 20 °C from ECB (2009) was selected for this risk
evaluation because out of the 9 remaining water solubility data, 6 data reported the water solubility from
ECB (2009).
F.1.5 Logarithmic OctanolrAir Partition Coefficient (log Koa)
Two high-quality log Koa data were identified through systematic review. Okeme et al. (2020) gave a
log Koa range of 7.85 to 7.93. Yaman et al. (2020) gave a log Koa value of 7.91. Because 7.91 is within
the range of 7.85 to 7.93, the Okeme et al. (2020) data was selected for use in the risk evaluation.
F.1.6 Henry's Law Constant
A Henry's Law constant of 2.55x 10~8 atm-m3/mol at 25 °C was reported in the final scope (U.S. EPA.
2020b). It was calculated using the Bond method in HENRYWIN™, which is an estimation method that
splits a compound into a summation of the individual bonds that comprise the compound (U.S. EPA.
2017a). However, when measured Henry's Law constant values are not available, a calculated value
based on high-quality measured water solubility and vapor pressure data are typically preferred over an
estimated value (Meylan and Howard. 1991). With a high-quality measured vapor pressure of 0.0613
mm Hg from Dobry and Keller (1957) and water solubility of 7,820 mg/L from ECB (2009). the revised
Henry's Law constant is 2,945x ] 0 6 atm-m3/mol at 25 °C. Systematic review identified two Henry's
Law constant data: one high-quality (Ekpe et al.. 2020) and one medium-quality data (IPCS. 1998). Both
data were not included in this risk evaluation because a calculated Henry's Law constant value based on
high-quality measured water solubility and vapor pressure data were available for use in the risk
evaluation.
F.1.7 Flash Point
Eight high-quality and four medium-quality flash point data were identified through systematic review.
The flash point data ranged from 200 to 252 °C. In general, flash point is measured using either an open
cup or closed cup technique. The closed cup technique normally gives lower values for the flash point
than open cup (approximately 5-10 °C lower). The extracted flash point data include values measured
Page 471 of 638
-------
using both closed cup and open cup techniques and some sources not reporting the technique used. Four
medium-quality data were excluded for this risk evaluation because high-quality flash point data are
available. The 216 °C datum extracted from U.S. EPA (2015a) and Lewis and Hawlev (2007) was
excluded because the analytical method was not provided and there was no indication that a reliable
method was used. The 202 °C datum extracted from IPCS (1998) was excluded because the data were
extracted from a secondary source without peer review and did not provide a reference of the original
source. The 200 °C datum extracted from U.S. EPA (2015a) was excluded because the test sample
appeared to catch fire at approximately 200 °C but did not show a distinct flash point as defined by the
ASTM D93 method. The 232 °C datum extracted from Toscano and Coleman (2012) and Sigma-Aldrich
(2019) was excluded because the analytical method used was not reported. Between the remaining two
high-quality flash point data, the 225 °C datum extracted from U.S. EPA (2015a) was selected for use in
this risk evaluation because flash point is defined as "the lowest temperature at which a chemical will
ignite with an ignition source."
F.1.8 Autoflammability
Three medium-quality autoflammability data were identified through systematic review. The 480 °C
datum extracted from ECB (2009) and ILO (2019) was selected for use in this risk evaluation because
autoflammability is defined as "the lowest temperature at which a chemical will spontaneously combust
without an ignition source." Therefore, the 1,115 °F (-602 °C) datum extracted from NTP (1992) was
excluded.
F.2 Fate and Transport
F.2.1 Approach and Methodology
EPA conducted a Tier I assessment to identify the environmental compartments (i.e., water, sediment,
biosolids, soil, groundwater, air) of major and minor relevance to the fate and transport of TCEP. Next, a
Tier II assessment was conducted to identify the fate pathways and media most likely to cause exposure
to environmental releases. Media-specific fate analyses were performed as described in Sections F.2.2,
F.2.3, and F.2.4.
F.2.1.1 EPI Suite™ Model Inputs
To set up EPI Suite™ for estimating fate properties of TCEP, the physical and chemical properties were
input based on the values in Table 2-1. EPI Suite™ was run using default settings (i.e., no other
parameters were changed or input) (Figure Apx F-2).
Page 472 of 638
-------
SJ EPI Suite
_Jn|xj
The Estimation Programs Interface (EPI) SuiteTM was developed by the US Environmental Protection Agency's Office of Pollution Prevention
and Toxics and Syracuse Research Corporation (SRC). It is a screening-level tool, intended for use in applications such as to quickly screen
chemicals for release potential and "bin" chemicals by priority for future work. Estimated values should not be used when experimental
(measured) values are available.
EPI SuiteTM cannot be used for all chemical substances. The intended application domain is organic chemicals. Inorganic and organometallic
chemicals generally are outside the domain.
Important information on the performance, development and application of EPI SuiteTM and the individual programs within it can be
found under the Help tab. Copyright 2000-2012 United States Environmental Protection Agency for EPI SuiteTM and all component
programs except BioHCWIN and KOAWIN.
FigureApx F-2. Screen Capture of EPI Suite1' Parameters Used to Calculate Fate and Physical
and Chemical Properties for TCEP
F.2.1.2 Fugacity Modeling
Because no current data were being reported to the TRI or DMR, TCEP releases to the environment
could not be estimated. The approach described by Mackav et al. (1996 using the Level III Fugacity
Model in EPI Suite™ (LEV3EPI1M) was used for TCEP's Tier II analysis. LEV3EPI™ is described as a
steady-state, non-equilibrium model that uses a chemical's physical and chemical properties and
degradation rates to predict partitioning of the chemical between environmental compartments and its
persistence in a model environment (U.S. EPA. 2017a). TCEP's physical and chemical properties were
taken directly from Table 2-1. Environmental release information is useful for fugacity modeling
because the emission rates will predict a real-time percent mass distribution for each medium. Instead,
environmental degradation half-lives were taken from high-quality studies that were identified through
systematic review to reduce levels of uncertainties. The results of the Level III Fugacity model reported
releases that show that the emissions of TCEP will primarily partition to water (59%) and soil (38%)
with less than 2 percent partitioning to sediment and less than 1 percent to air (Figure Apx F-3). These
results reiterate the Tier I analysis results that water and soil are expected to be an important
environmental media for TCEP released to the environment. The LEV3EPI™ results were consistent
with environmental monitoring data. Further discussion of TCEP partitioning can be found in Sections
F.2.2, F.2.3, and F.2.4.
Page 473 of 638
-------
100%
100% Soil Release 100% Air Release 100% Water Release Equal Releases
Air ¦ Water BSoil I Sediment
FigureApx F-3. EPI Suite™ Level III Fugacity Modeling Graphical Result for TCEP
F.2.1.3 OECD Pov and LRTP Screening Tool
TCEP's long-range transport potential (LRTP) was evaluated by using OECD's Overall Environmental
Persistence (Pov) and LRTP Screening Tool, Version 2.2 (Wegmann et al.. 2009). The OECD POV and
LRTP Tool is in a spreadsheet format containing multimedia chemical fate models that were designed
based on the recommendations of the OECD expert group to estimate environmental persistence and
LRTP of organic chemicals at a screening level. With a chemical's physical and chemical properties, the
OECD POV and LRTP Tool will be able to predict its Pov, characteristic travel distance (CTD), and
transfer efficiency (TE). Pov is the overall persistence in the whole environment in days, CTD quantifies
the distance in kilometers (km) from the point of release to the point at which the concentration has
dropped to 1/e, or approximately 37 percent of its initial value, and TE estimates the percentage of
emitted chemical that is deposited to surface media after transport away from the region of release. The
OECD Pov and LRTP Screening Tool calculates two emission scenarios specific CTD values, for
emissions to air and water. Only transport in the medium that receives the emission is considered, thus
CTD in air is calculated from the emission-to-air scenario and CTD in water is calculated from the
emission-to-water scenario. No CTD is calculated for emissions to soil because soil is not considered to
be mobile. The physical and chemical properties were input based on the values in Table 2-1 and Table
2-2 (Figure_Apx F-4). The modeling results will be discussed further in Sections F.2.2 and F.2.3.1.
Page 474 of 638
-------
OECD POV & LRTP
Screening Tool*
Main Menu
Help
Preferences
Select chemicals to evaluate
Simultaneous runs of one database and one chemical are possible.
Databases
Reference Chemicals
Generic PCB Hornoloques
A
V
Deselect | Manage DB» |
Database Status:
~
Single Chemical
Monte Carlo Parameters
Name
TCEP
Dispersion factors
Molecular mass
285.49
for each property
Log Ka,
-3.919
~
~~5 *
Log KaK
1.78
~
5 j
Half life in air (h)
5.80E+00
~
10
Half life in water (h)
1.00E+04
~
10
Half life in soil (h)
4.25E+02
~
10
Clear
1
Reset
Chemical Status:
~
Calculate
R Include Monte Carlo Analysis for Single Chemical
Color Codes
H
Results already present
~
No warnings: calculation possible
~
Warnings: calculation still possible
~
Errors: calculation impossible
~
No data entered
' A manual describing this software is provided on the Help page.
FigureApx F-4. Screen Capture of OECD Pov and LRTP Screening Tool Parameters Used to
Calculate TCEP's LRTP
F.2.1.4 Evidence Integration
A brief description of evidence integration for fate and transport is available in the 2021 Draft
Systematic Review Protocol (U.S. EPA. 2021a). Additional details on fate and transport evidence
integration are provided here.
The environmental fate characteristics given in Appendix C of the Final Scope of the Risk Evaluation
for Tris(2-chloroethyl) Phosphate (TCEP) CASRN115-96-8 (U.S. EPA. 2020b) were identified prior to
completing the systematic review. The following sections summarize the findings and provide the
rationale for selecting the environmental fate characteristics that was given in Table 2-2.
F.2.2 Air and Atmosphere
TCEP in its pure form is a liquid at environmental temperatures with a melting point of-55 °C (DOE.
2016; U.S. EPA. 2015a. b; Toscano and Coleman. 2012) and a vapor pressure of 0.0613 mm Hg at 25
°C (U.S. EPA. 2019b; Dobry and Keller. 1957). The log Koa range of 7.5 to 7.98 indicates that TCEP is
Page 475 of 638
-------
expected to adsorb to the organic carbon present in airborne particles (Okeme et al.. 2020; Ji et al.. 2019;
Wang etal.. 2017b).
As an SVOC, TCEP will exist in both the gas and particle phases (Wang et al.. 2020a; Okeme. 2018;
TERA. 2015). Results from air monitoring studies reported concentrations of gaseous TCEP up to 6,499
pg/m3 (Ma et al.. 2021; Wu et al.. 2020) and particle bound TCEP up to 2,100 pg/m3 in North America
(Wu et al.. 2020; Abdollahi et al.. 2017; Salamova et al.. 2016; Salamova et al.. 2014; Shoeib et al..
2014). Multiple studies have identified urban sources as sources of TCEP in the environment through
fugitive emissions to air (Abdollahi et al.. 2017; Luo et al.. 2015; Moller et al.. 2011). Although the
exact sources of TCEP emissions from urban environment are unknown, they are likely the articles that
were treated with or containing TCEP (Abdollahi et al.. 2017; Luo et al.. 2015; Wei et al.. 2014; Moller
et al.. 2011; Aston et al.. 1996).
Compared to outdoor air, TCEP concentrations are significantly higher in indoor air because TCEP has
the potential to volatilize from treated products and diffuse into air, as well as partition onto dust due to
its use as an additive (Oi et al.. 2019; TERA 2015; Liu et al.. 2014; AT SDR. 2012; EC. 2009; NICNAS.
2001). In northern California, indoor air concentrations of TCEP were detected up to 15,340 pg/m3
(Bradman et al.. 2014) and dust concentrations was measured up to 6.84 |ig/g (Bradman et al.. 2012). In
addition, TCEP is a known impurity in 2,2-bis(chloromethyl)-propane-l,3-diyltetrakis(2-chloroethyl)
bisphosphate (V6) commercial mixtures that are primarily used in furniture and automobile foam.
Higher concentrations of TCEP (up to 50.12 |ig/g) were found in dust samples that were collected from
the surfaces of the front and back seats of automobiles in Boston, MA (Fang et al.. 2013).
TCEP is not expected to undergo significant direct photolysis in the atmosphere because its chemical
structure does not absorb light at wavelengths greater than 290 nm (HSDB. 2015). TCEP in the gaseous
phase is expected to degrade rapidly by reaction with photochemically produced hydroxyl radicals
(•OH) in the atmosphere. A half-life of 5.8 hours was calculated from the AOPWIN™ module in EPI
Suite™ using an estimated rate constant of 2.2xl0~u cm3/molecules-second at 25 °C, assuming an
atmospheric hydroxyl radical concentration of 1.5><106 molecules/cm3 and a 12-hour day (U.S. EPA.
2017a). The atmospheric half-life of TCEP does not pertain to indoor environments due to lower
hydroxyl radical concentrations, less mixing of air, and lower sunlight intensity.
TCEP has been detected in air and snow in remote locations such as the Arctic and Antarctica (Na et al..
2020; Wang et al.. 2020a; Xie et al.. 2020; Rauert et al.. 2018; Li et al.. 2017b; Stihring et al.. 2016;
Cheng et al.. 2013b; Moller et al.. 2012; NIVA. 2008). Particle-bound TCEP was found to be highly
persistent in the atmosphere and had slower rates for the reaction with hydroxyl radicals due to the
presence of atmospheric water (Wu et al.. 2020; Li et al.. 2017a; Liu et al.. 2014). Particle-bound TCEP
is primarily removed from the atmosphere by wet or dry deposition. Based on its physical and chemical
properties and short half-life in the atmosphere (ti/2 = 5.8 hours), TCEP was assumed to be not persistent
in the air (U.S. EPA. 2017a). The OECD Pov and LRTP Screening Tool was run to get additional
information on TCEP's long-range transport potential in the air. For TCEP emissions in air, a Pov of 11
days, CTD of 118 km (-73 miles), and TE of 0.0142 percent were given using a molecular mass of
285.49 g/mol, logarithmic air:water partitioning coefficient (log Kaw) of-3.919, and logarithmic
octanol:water partitioning coefficient (log Kow) of 1.78 along with atmospheric half-life of 5.8 hours,
water half-life of 10,000 hours, and soil half-life of 424.8 hours (Figure Apx F-4). A CTD of 118 km
(-73 miles) suggests that TCEP does not have the potential to undergo long-range transport in the air
and a TE of 0.0142 percent suggests that negligible fraction of TCEP emitted to air will be deposited to
surface media such as water. CTD can also be calculated using the LEV3EPI™ module in EPI Suite™
without considerations for advection (U.S. EPA. 2017a; Beyer et al.. 2000). After entering TCEP's
Page 476 of 638
-------
physical and chemical properties (Figure Apx F-2), a CTD of 238 km (-148 miles) was calculated.
Particle-bound TCEP has the potential to undergo long-range atmospheric transport (LRAT) and it is
likely the reason why TCEP is found in the Arctic and other remote locations with no source of releases.
TCEP's LRTP could be crucially underestimated when using gaseous phase atmospheric half-life in
multimedia models like the OECD Pov and LRTP Screening Tool.
F.2.3 Aquatic Environments
Wastewater treatment effluent, atmospheric deposition, air-water gaseous exchange, and runoff have
been identified as sources of TCEP detected in aquatic and marine environments, especially in urban
areas (Ma et al.. 2021; Cristate et al.. 2019; Guo et al.. 2017a; Kim et al.. 2017).
F.2.3.1 Surface Water
TCEP is not expected to undergo abiotic degradation processes such as hydrolysis and photolysis in
aquatic environments under environmentally relevant conditions. The rate of hydrolysis will be highly
dependent on pH and temperature. TCEP showed no significant hydrolysis over 35 days at pH levels of
7, 9, and 11 at 20 °C, but an extensive degradation occurred when the pH level was adjusted to 13 (ti/2 =
0.083 days) (Su et al.. 2016). A hydrolysis study by Saint-Hilaire et al. (2011) observed the pH-
dependent hydrolysis of TCEP between pH 8 to 13 at 50 °C and confirmed that TCEP's hydrolysis rates
increased as pH levels increased. TCEP's hydrolysis half-life was estimated to be approximately 2 years
at pH level of 8 at 25 °C. In addition, TCEP's hydrolysis rates also increased in the presence of reduced
sulfur species. The calculated half-lives for TCEP after reacting with 5.6 mM bisulfide (HS~) and 0.33
mM polysulfides (S „") were 90 and 30 days, respectively. The results also indicated that the three
reduced sulfur species reacted with TCEP in a nucleophilic substitution reaction with bis(chloroethyl)
phosphate (BCEP) being the major transformation product. The hydrolysis half-lives estimated by
QSAR models were found to be inconsistent with experimental values. HYDROWIN™, an aqueous
hydrolysis rate program in EPI Suite™, estimated TCEP's half-life to be approximately 20 days at pH 5
to 9 and approximately 17 days at pH 10 (U.S. EPA. 2017a). However, the half-life values from
HYDROWIN™ were not included in this risk evaluation because the half-life values from high-quality
hydrolysis studies mentioned above are available. In addition, it is unlikely for TCEP undergo indirect
photolysis. No photolytic degradation was observed after exposing TCEP to natural sunlight for 15 days
in lake water (Regnery and Ptittmann. 2010a). Other experimental studies also observed no photolytic
degradation (Chen et al.. 2019b; Lee et al.. 2014; Watts and Linden. 2009. 2008).
For biotic degradation in water, TCEP is not readily biodegradable under aerobic conditions. In a ready
biodegradability test using the Modified Sturm test (OECD TG 301B), TCEP showed a minimal
degradation after 28 days and the cumulative carbon dioxide (CO2) production was negligible (Life
Sciences Research Ltd. 1990b). In another ready biodegradability test using the Closed Bottle test
(OECD TG 301D), TCEP was not readily biodegradable (Life Sciences Research Ltd. 1990c). Based on
these two biodegradation studies, rapid biodegradation of TCEP is not likely when it is released to
surface water.
A limited number of test results on anaerobic biodegradability of TCEP were available. Kawagoshi et al.
(2002) reported that TCEP did not undergo biodegradation under anaerobic condition after 60 days
using leachate from a sea-based solid waste disposal site in Japan. This study was not selected for use in
the risk evaluation because it was rated as a medium-quality study because critical information on test
conditions was not included and there was insufficient evidence to confirm that TCEP disappearance
was not likely due to other processes. Due to lack of anaerobic biodegradation studies on TCEP, no
anaerobic biodegradation data were selected for this risk evaluation.
Page 477 of 638
-------
Two studies showed that TCEP was able to undergo volatilization from oceans and had the highest
water-to-air emission flux in two monitoring studies. In Li et al. (2017bI TCEP volatilization from
seawater to air was seen in all samples across the North Atlantic and the Arctic, and equilibrium was
reached in some samples that was caused by relatively low TCEP concentrations in seawater. A similar
result was seen in another air-water gaseous exchange study on a coastal site where TCEP had the
highest emission flux in water (Wang et al.. 2018b). Both studies suggest that the air-water gaseous
exchange is an important process for TCEP to transport between the air and water, causing a secondary
pollution. TCEP's volatilization behavior did not align with its physical and chemical properties and
modeling prediction. A Henry's Law constant of 2.945x 10~6 atmm3/mol at 25 °C (Table 2-1) indicates
that TCEP is not expected to volatilize from surface water (TERA. 2015; Toscano and Coleman. 2012;
Regnery and Puettmann. 2009; Dobry and Keller. 1957). A Henry's Law constant is equivalent to aKAW
of 1.21 x 10~4 or log Kaw of -3.19 at 25 °C, which indicates that TCEP will favor water over air (U.S.
EPA. 2017a). However, monitoring studies disproved that assumption and suggested that volatilization
is significant. The Water Volatilization Program in EPI Suite™ estimated the volatilization half-lives of
TCEP from a model river and lake and default settings were applied (see default settings in FigureApx
F-2). TCEP's volatilization half-life from a model river was 337.6 hours (-14 days), and 3,825 hours
(-159 days) for the model lake (U.S. EPA. 2017a). Overall, TCEP's potential to volatilize from water
can be underestimated significantly if one relies solely on interpreting its physical and chemical
properties or using QSAR models. Only experimental data would properly describe TCEP's
volatilization behavior.
When precipitation events occur, TCEP's mobility in the environment will be greatly enhanced because
rain and snow are believed to be effective scavengers of organic contaminants (Awonaike et al.. 2021;
Mihailovic and Fries. 2012; Regnery and Puettmann. 2009; Lei and Wania. 2004). Atmospheric
deposition has been identified as an important source of TCEP to surface water, especially in urban
areas. Several studies showed that higher TCEP concentrations in precipitation were generally seen in
densely populated areas with high traffic volume (Kim and Kannan. 2018; Regnery and Piittmann.
2010b; Regnery and Puettmann. 2009; Marklund et al.. 2005b). In addition, storm water and urban
runoff can contribute to additional emissions to surface water. The presence of TCEP in runoffs can be
attributed to TCEP's use as an additive in car interiors and building materials and high water solubility.
During periods without precipitation events, dry deposition is expected to occur (Na et al.. 2020; Li et
al.. 2017b; Lai et al.. 2015; Mihailovic and Fries. 2012).
The OECD Pov and LRTP Screening Tool was run to get additional information on TCEP's LRTP in
water (Figure Apx F-4). For TCEP emissions in water, a Pov of 414 days, CTD of 707 km (-439.3
miles), and TE of 0.0014 percent were estimated. A CTD of 707 km suggests that TCEP does not have
the potential to undergo long-range transport. Yet, TCEP was detected in the waters of the Arctic, which
is approximately 1,775 miles away from New York City (Na et al.. 2020; McDonough et al.. 2018; Li et
al.. 2017b). As previously mentioned, snow is an effective scavenger of organic contaminants, and it is
possible to see the TCEP concentration in adjacent surface water spike from global warming. In
addition, plastic debris, and ocean currents (e.g., gyres) may have played a role in TCEP being widely
distributed in aquatic and marine environments (Xie et al.. 2020; Li et al.. 2017b; Cheng et al.. 2013a;
Andresen et al.. 2007). Plastic debris existing in marine environments have been found to contain
various types of chemicals (Takada and Karapanagioti. 2019; Zhang et al.. 2018a; Mato et al.. 2001).
Plastic products typically contain various additives that are used at high volume fractions in the plastic
formulation such as plasticizers and flame retardants to maintain their performances (Takada and
Karapanagioti. 2019). In locations where waste is uncollected or unmanaged, plastic wastes are likely to
end up as litter where TCEP are released into the open environment. Extreme events such as storms,
floods, cyclones, tidal waves, and tsunamis, are also a significant immediate source of land-based plastic
Page 478 of 638
-------
debris. Plastic wastes containing TCEP can potentially migrate from the plastic product to water by the
weathering of microplastics (Hahladakis et al.. 2018). Because TCEP has primarily been used as an
additive flame retardant and plasticizers, they can easily leach from plastic wastes. Furthermore, plastic
debris (e.g., macroplastics, microplastics) could act as carriers for TCEP. The high specific surface areas
of microplastics make them a good sorbent for hydrophobic and hydrophilic organic chemicals (Zhang
et al.. 2018a). Widely used plastics such as polyvinyl chloride (PVC) and polyethylene (PE) sorb
organic pollutants from seawater after they are exposed to environmental conditions (Takada and
Karapanagioti. 2019). In Chen et al. (2019a) TCEP was seen to sorb onto PVC and PE microplastics in
seawater. When the temperature was in the range of 5 to 15 °C, the adsorption capacity of TCEP
increased with increasing temperature, but when the temperature was greater than 15 °C, the adsorption
capacity decreased with increasing temperature. Through adsorbing pollutants from surrounding
seawater, microplastics can accumulate and increase the concentrations of pollutants up to the order of
106 (Mato et al.. 2001). Plastic wastes are found in the ocean all over the world and they can travel long
distances, especially to remote regions.
Based on the findings provided above, TCEP has the potential undergo long-range transport in water and
its LRTP could be underestimated when using multimedia models like the OECD Pov and LRTP
Screening Tool.
F.2.3.2 Sediments
TCEP can be transported to sediment from overlying surface water by advection and dispersion of
dissolved TCEP and by deposition of suspended solids containing TCEP. TCEP is expected to partition
to organic matter in suspended and benthic sediments based on its measured sediment logarithmic
organic carbon:water partition coefficient (log Koc) values ranging from 3.23 to 3.46 (Zhang et al..
2021; Wang et al.. 2018a; Zhang et al.. 2018b). Higher concentrations of TCEP in sediment are expected
to be found at potential source locations (e.g., near urban and industrialized areas) (Chokwe and
Okonkwo. 2019; Tan et al.. 2019; Lee et al.. 2018; Wang et al.. 2018a; Cao et al.. 2017; Maruva et al..
2016; Cristate et al.. 2013).
A limited number of test results on anaerobic biodegradability of TCEP were available (see Appendix
F.2.3.1). Systematic review did not identify anaerobic biodegradation studies for TCEP. However,
systematic review identified two published risk assessments that reported an anaerobic biodegradation
study. EC (2009) and U.S. EPA (2015a) reported that no degradation was observed for TCEP in an
anaerobic biodegradation study after 58 days using ISO 11734, which is equivalent to OECD TG 311
(Noack. 1993). This result was not selected for use in the risk evaluation because the original study by
Noack (1993) was published in German; therefore, it did not undergo the systematic review process.
After systematic review concluded, four studies observing anaerobic biodegradation of TCEP were
identified. Pang et al. (2018) studied the fate of TCEP in sewage sludge with four different treatments,
two of which were anaerobic digestion studies (Treatments 3 and 4). The results of this study confirm
the recalcitrance of TCEP under anaerobic conditions and TCEP concentrations in both treatments were
seen to increase over time. The removal rates were -0.41 and -74.8 percent for Treatments 3 and 4,
respectively. The authors did not discuss the reasons for the increase in TCEP's final concentrations, but
the authors did note that the final concentrations of Tri(n-butyl) phosphate (TnBP) increased during
composting due to the "inhomogeneity" of the composts. Because of this uncertainty, Pang et al. (2018)
is used qualitatively to support TCEP's persistence in anaerobic environment and will not be used to
derive a biodegradation half-life. Yang et al. (2021) carried out biodegradation experiments using
activated sludges from aerobic and anerobic ponds of the Nanjing Chendong STP in China mixed with
kitchen garbage biomass and agricultural residues. The biodegradation of TCEP under anaerobic
Page 479 of 638
-------
conditions was observed to be very slow. The removal of TCEP did not reach 80 percent after 60 days.
Because the media and inoculum used in Yang et al. (2021) comprised a rich nutrient profile and high
microbial biomass, the direct use of the reported half-lives for estimating persistence in environmental
sediments is not appropriate; therefore, this study was excluded from use in this risk evaluation. Yang et
al. (2023) studied the degradation of TCEP using an anaerobic enrichment culture from end-of-life
vehicles dismantling sites in Guangzhou, China. The study demonstrated that microbes, specifically
Dehcilococcoides sp., was able to anaerobically transform TCEP. However, the microbial community
was highly enriched and the system highly controlled. This study will not appropriately represent
TCEP's fate under anaerobic conditions and will not be considered in this risk evaluation.
When considering the results of the Pang et al. (2018) and Life Sciences Research Ltd (1990b) studies
(Table 2-2), it is highly likely that TCEP will not degrade under anaerobic conditions and be persistent
in the sediment compartment.
The rate of biodegradation in sediments can be estimated by extrapolation from aerobic biodegradation
testing or estimated by considering that the rate of anaerobic degradation is typically at least three to
four times slower (64 FR 60197. November 9, 1999). For the water compartment, TCEP did not pass a
ready biodegradability test (OECD TG 301B) (Life Sciences Research Ltd. 1990b) (Table 2-2), so a
default water half-life of 10,000 hours was used as described on page 4 in the Interim Guidance for
Using Ready and Inherent Biodegradability Tests to Derive Input Data for Multimedia Models and
Wastewater Treatment Plants (WWT) Models (9/1/2000) (U.S. EPA. 2000a). Considering that the rate of
anaerobic degradation is 3 to 4 times slower than aerobic biodegradation, the estimated half-life of
TCEP would be 30,000 to 40,000 hours in the sediment compartment.
F.2.3.3 Key Sources of Uncertainty
Several studies reported that TCEP concentrations in overlying water were higher when compared to
sediment (Lee et al.. 2018; Ma et al.. 2017; Brandsma et al.. 2015; Cao et al.. 2012; Kawagoshi et al..
1999). For example, Lee et al. (2018) reported a TCEP mean concentration of 255 ng/L in the waters of
Lake Shihwa in Korea, while the sediment had a mean TCEP concentration of 18.4 ng/g dry weight.
TCEP concentrations in surface water and sediment may vary in locations due to the competing
processes of advection, turbulence (sediment-water and air-water mixing), the bioturbation of benthic
animals, and resuspension of particulate matters. In addition, sediment concentrations were observed to
be correlated with sediment organic carbon content (Lee et al.. 2018; Wang et al.. 2018a; Xing et al..
2018; Zhong et al.. 2018) and TCEP's water solubility may be a major factor.
F.2.4 Terrestrial Environments
TCEP is released to terrestrial environments via land application of biosolids, disposal of solid waste to
landfills, and atmospheric deposition.
F.2.4.1 Soil
Based on its measured log Koc values in soil ranging from 2.08 to 2.52 (Table 2-2), TCEP accumulation
in soil is expected to be unlikely and TCEP may be mobile and eventually migrate to groundwater (see
Section F.2.4.2). TCEP in the soil was seen to be vertically transported to deeper soil horizons, causing
TCEP concentration in the surface soil to be lower (He et al.. 2017; Bacaloni et al.. 2008). Zhang et al.
(2022) reported that higher levels of TCEP was found deeper in the soil (30 to 80 cm) compared to the
surface soil samples (0 to 20 cm). Mihailovic and Fries (2012) reported a similar result in its study.
Page 480 of 638
-------
The estimated log Koc value for TCEP is 2.59, using the molecular connectivity index (MCI) method in
KOCWIN™ (U.S. EPA. 2017a). The estimated value from EPI Suite™ is not included in this risk
evaluation because the log Koc values from high-quality field studies are available.
Systematic review identified two high-quality studies on TCEP degradation in soil. Hurtado et al. (2017)
studied the degradation of TCEP in an agricultural soil from Spain. The soil had a sandy texture (90
percent sand, 8 percent silt, and 2 percent clay) and a total organic carbon content of 5 g/kg. After 40
days, 78 percent of TCEP degraded under aerobic conditions at test substance concentration of 50 |ig/kg.
A half-life of 17.7 days (Table 2-2) was estimated based on second-order kinetics. Another soil
degradation study was identified, but this study was evaluated as low-quality ((ECB. 2009). citing
(Brodsky et al.. 1997)). The primary degradation of TCEP at a concentration of 5 mg/kg soil was
conducted in a laboratory test system with standard soil for 100 days. The degradation kinetic curve was
fitted to a 2nd order square root function resulting in a DT50 of 167 days and DT90 of >100 days. In
addition, TCEP was seen to be slightly mobile in a leaching test. However, this study is not included in
this risk evaluation because the testing conditions, inoculum information, sampling and analytical
methods were not reported, and the omissions likely had an impact on the study results. Zhu et al. (2023)
characterized the attenuation of TCEP in landfill humus and subsoil taken from a non-sanitary landfill
that once was remediated. The study highlighted the roles of both abiotic degradation and biotic
degradation contributing to the reduction of TCEP in both soil types. The authors indicate possible co-
factors mediating the dechlorination of TCEP (e.g., iron, sulfur, and carbon levels) pointing to the
variety of conditions under which TCEP may be more readily degraded. Given the specificity of the
experiments, Zhu et al. (2023) is not appropriate for deriving anaerobic biodegradation rates in natural
environments.
TCEP in soil can re-volatilize from contaminated soil into the atmosphere causing a secondary pollution.
A Henry's Law constant of 2.945x 10~6 atm-m3/mol at 25 °C, calculated based on a vapor pressure of
0.0613 mm Hg and a water solubility of 7,820 mg/L at 25 °C, indicates that TCEP is not expected to
volatilize from dry soil but possibly from moist soil (ATSDR. 2012; Toscano and Coleman. 2012;
Regnery and Puettmann. 2009; Dobry and Keller. 1957). Yet, there are field studies showing that TCEP
underwent an air-soil exchange. In Wang et al. (2020b). the air-soil exchange behavior of TCEP varied
between locations. TCEP was observed to be at an air-soil exchange equilibrium in the suburban and
rural areas, but net volatilization occurred in the urban area. The highest volatilization flux was found at
a site near a bus terminal. Yadav et al. (2018) reported net volatilization from soil to the air as TCEP's
principal process in air-soil exchange. Han et al. (2020) reported a net volatilization in a sampling site
located in the Arctic.
Also, several studies have reported that atmospheric deposition of TCEP may have contributed to soil
contamination because there were no point sources nearby (Ji et al.. 2019; Ren et al.. 2019; Fries and
Mihailovic. 2011; Mihailovic et al.. 2011). In Bacaloni et al. (2008). lake water samples were collected
from three remote volcanic lakes in central Italy. The three lakes were specifically chosen because there
were no local contamination sources (e.g., tributaries, industries, sewage treatment plants) nearby.
Therefore, the possible sources of contamination would be from local anthropogenic activities, long-
range transport and deposition from rainfall, or runoff processes. TCEP was detected in all three lakes at
the ng/L level and the maximum concentrations occurred during the late summer to autumn months
(August to October), which coincides with higher tourism activity and vehicular traffic at all three
locations. In Han et al. (2020). the net deposition from air to soil was found to be predominant in four
out of five sampling sites in the Arctic.
Page 481 of 638
-------
F.2.4.2 Groundwater
There are two sources of TCEP in the environment that may contaminate groundwaters. Point sources
include wastewater effluents and landfill leachates and are discussed in Sections F.2.5.2 and F.2.4.3.
Diffuse sources include storm water runoff and runoff from biosolids applied to agricultural land and are
discussed in sections F.2.3.1 and F.2.4.4.
Municipal solid waste landfills (MSWLFs) can be a source of TCEP groundwater contamination.
Historic landfills are more likely to lack the infrastructure of modern landfills, such as liners, leachate
collection systems, and reactive barriers, which would prevent leachate from entering the groundwater
system (Propp et al.. 2021; Lapworth et al.. 2012; Barnes et al.. 2004).
Propp et al. (2021) assessed contaminants of emerging concern in leachate-impacted groundwater from
20 closed MSWLFs in Ontario, Canada. Those "historic" landfills had been closed for at least three
decades. High concentrations of TCEP were reported in groundwater up to 2.92 |ig/L. In addition,
Buszka et al. (2009) collected groundwater samples from a domestic well located in a neighborhood east
of the Himco Dump, which is an unlined landfill that was used for commercial, industrial, medical, and
general waste disposal from 1960 to 1976 in Elkhart, Indiana. TCEP concentration ranged from 0.65 to
0.74 |ig/L. Both studies suggests that TCEP in landfill impacted groundwater was resistant to biotic and
abiotic degradation processes and is very persistent. Barnes et al. (2004) collected groundwater samples
from a historic landfill in central Oklahoma. The landfill was unlined and built adjacent to the Canadian
River in 1920, then covered with a clay cap and vegetated when it was permanently closed in 1985.
TCEP concentration of 0.36 |ig/L was measured in a well that was 3.28 feet away from the landfill.
However, a TCEP concentration of 0.74 |ig/L was measured in a well located 305 feet away from the
landfill. This shows TCEP's potential to be transported away from point sources and enter groundwater.
F.2.4.3 Landfills
TCEP is not considered a hazardous waste, so it is not listed under Subtitle C of the Resource
Conservation and Recovery Act (RCRA) (40 CFR 261). Solid waste containing TCEP can be disposed
in MSWLFs or industrial waste landfills (i.e., construction and demolition [C&D] debris landfills).
MSWLFs that were built after 1991 are required to use a composite liner and a leachate collection
system. The composite liner includes a minimum of 30-mil flexible membrane liner (FML) overlaying a
two-foot layer of compacted soil lining the bottom and sides of the landfill (40 CFR 258.40). It is
expected that solid waste containing TCEP will be disposed to a lined landfill with a leachate collection
system. However, historic landfills are likely to lack the infrastructure of modern landfills, such as
liners, leachate collection systems, and reactive barriers (Propp et al.. 2021; Lapworth et al.. 2012;
Barnes et al.. 2004). Leachate-impacted groundwater in historic landfills is discussed in Section F.2.4.2.
As discussed in Section 2.2.2, TCEP is primarily used as an additive plasticizer and flame retardant.
When used as an additive, TCEP is added to manufactured materials via physical mixing rather than
chemical bonding (Oi et al.. 2019; Liu et al.. 2014; AT SDR. 2012; EC. 2009; NICNAS. 2001).
Consequently, it is highly likely that TCEP will be released from the solid wastes and enter the leachate.
Leachates from 11 landfill sites in Japan reported TCEP concentrations in the range of 6 to 30,100 ng/L
(Yasuhara et al.. 1999). The maximum concentration of TCEP was reported in a landfill that consisted
of waste plastics, waste combustion residue, plants, and domestic incombustible wastes. Several other
studies also showed high concentrations of TCEP in leachate samples collected from MSWLFs in the
United States and China (Oi et al.. 2019; Deng et al.. 2018a; Masoner et al.. 2016; Masoner et al..
2014b).
Page 482 of 638
-------
Landfill leachate can be discharged to WWTPs and the release of TCEP to surface water from treated
landfill leachate will depend on the removal of TCEP during wastewater treatment (see Section F.2.5.2).
The fate and transport of TCEP entering surface water is discussed in Section F.2.3.1.
F.2.4.4 Biosolids
Sludge is defined as the solid, semi-solid, or liquid residue generated by wastewater treatment processes.
The term "biosolids" refers to treated sludge that meet the EPA pollutant and pathogen requirements for
land application and surface disposal (40 CFR 503).
Because TCEP is resistant to degradation in wastewater treatment, some residual concentrations of
TCEP may be present in biosolids and transferred to surface soil during land application. TCEP
concentrations between 78.9 to 317 ng/g dry weight were detected in sewage sludge collected from
wastewater treatment plants located in the United States (Wang et al.. 2019c; Kim et al.. 2017). An
anaerobic digestion study using sewage sludge showed that TCEP was persistent under anaerobic
conditions (Pang et al.. 2018). It is likely that dissolved TCEP will eventually reach surface water via
runoff after the land application of biosolids due to its persistence.
F.2.4.5 Key Sources of Uncertainty
There are significant differences between the predicted and the field observed log Koc values. The
predicted log Koc values are generally lower than the ones reported from field studies. The log Koc
reported in previous assessments of TCEP were in the range of 2.04 to 2.59 (TERA. 2015; AT SDR.
2012; EC. 2009; ECB. 2009; NICNAS. 2001). Koc values within this range are associated with low
sorption to soil and will be able to migrate to groundwater. However, a range of 2.5 to 4.3 was obtained
from several field studies (Awonaike et al.. 2021; Zhang et al.. 2021; Wang et al.. 2018a; Zhang et al..
2018b). Log Koc within this range are associated with moderate to strong sorption to soil, sediment, and
suspended solids.
F.2.5 Persistence Potential
Biotic and abiotic degradation studies have shown TCEP to be persistent. In the atmosphere, TCEP in
the gaseous phase will be degraded by reacting with hydroxyl radicals (*OH), but particle-phase TCEP
will not be degraded (see Section F.2.2). TCEP does not undergo hydrolysis under environmentally
relevant conditions and is persistent in water (see Section F.2.3.1), sediment (see Section F.2.3.2), and
soil (see Section F.2.4.1). Using the Level III Fugacity model in EPI Suite™ (LEV3EPI™) (see Section
F.2.1.2), TCEP's overall environmental half-life was estimated to be approximately 168 days (U.S.
EPA. 2017a). Therefore, TCEP is expected to be persistent in the atmosphere as well as aquatic and
terrestrial environments.
F.2.5.1 Destruction and Removal Efficiency
Destruction and removal efficiency is a percentage that represents the mass of a pollutant removed or
destroyed in a thermal incinerator in relative to the mass that entered the system. EPA requires that
hazardous waste incineration systems destroy and remove at least 99.99 percent of each harmful
chemical in the waste, including treated hazardous waste (46 FR 7684. January 23, 1981).
Only one study was identified in regard to thermal treatment and open burning of articles containing
TCEP. Li et al. (2019a) reported that the articles released TCEP in the range of 9,800 to 49,000 ng/g
after undergoing thermal treatment at 300 °C for 150 minutes. For open burning, the articles released
TCEP in the range of 1,000 to 2,600 ng/g after being exposed to an open flame for 3 minutes at 800 to
1,350 °C. These results showed that TCEP was not completely destroyed. This finding, however, is
Page 483 of 638
-------
expected because flame retardant-containing materials are known/intended to have reduced
flammability, which can result in incomplete combustion.
When undergoing thermal degradation in air at 220 °C and higher, TCEP will rapidly decompose to
produce numerous toxic byproducts, including 1,2-dichloroethane (C2H4CI2), vinyl chloride (C2H3CI),
hydrogen chloride (HC1), carbon monoxide (CO), and acetaldehyde (C2H4O), among others (U.S. EPA.
2015a: NICNAS. 2001; Muir. 1984; Paciorek et al.. 1978).
Because open burning can contribute to the emission of TCEP or other toxic byproducts to the
surrounding environment (Matsukami et al.. 2015). thermal treatment and open burning are not
favorable options for the disposal of TCEP.
F.2.5.2 Removal in Wastewater
Wastewater treatment is performed to remove contaminants from wastewater using physical, biological,
and chemical processes. Generally, municipal wastewater treatment facilities apply primary and
secondary treatments. During the primary treatment, screens, grit chambers, and settling tanks are used
to remove solids from wastewater. After undergoing primary treatment, the wastewater typically
undergoes a secondary treatment. Secondary treatment processes can remove up to 90 percent of the
organic matter in wastewater using biological treatment processes such as trickling filters or activated
sludge. Sometimes an additional stage of treatment, such as tertiary treatment, is utilized to further clean
water prior to release using advanced treatment techniques (e.g., ozonation). A negative removal
efficiency can be reported if the pollutant concentration is higher in the effluents than the pollutant
concentration in the influents.
Because TCEP is not readily biodegradable under aerobic conditions based on two ready
biodegradability tests (Life Sciences Research Ltd. 1990b. c), it is not expected to be removed from
wastewater by biodegradation. This conclusion is supported by STPWIN™, an EPI Suite™ module that
estimates chemical removal in sewage treatment plants. STPWIN™ estimated that a total of 2.23 percent
of TCEP in wastewater will be removed: 0.08 percent by biodegradation, 0.17 percent by air stripping,
and 1.99 percent by sorption to sludge (U.S. EPA. 2017a). STPWIN™ simulates a conventional
wastewater treatment plant that uses activated sludge secondary treatment. The biodegradation half-life
parameter was set to 10,000 hours for the primary clarifier, aeration vessel, and settling tank, which is a
default for recalcitrant chemicals. The physical and chemical properties for TCEP given in Table 2-1
were used (Figure Apx F-2). The results from STPWIN™ were not included in this risk evaluation
because high-quality wastewater treatment studies are available.
A total of 19 wastewater treatment studies were identified during systematic review. Seven were
evaluated and rated as medium-quality studies. These studies were not included in this risk evaluation.
Numerous high-quality wastewater treatment studies reported either a negative removal efficiency or a
removal of less than 10 percent for TCEP after undergoing primary and secondary treatments. An
overall TCEP removal of-60.2 percent was calculated for a municipal wastewater treatment in
Frankfurt, Germany (Fries and Puttmann. 2001). An average overall TCEP removal of-32.2 percent
was calculated from the removals reported for five activated sludge treatment plants in Catalonia, Spain
(Cristate et al.. 2016).
An TCEP removal of-18.9 percent removal was calculated for a municipal wastewater treatment plant
in Beijing, China (Liang and Liu. 2016). TCEP was not removed (0%) in two activated sludge treatment
plants in western Germany (Meyer and Bester. 2004) and an activated sludge treatment plant in South
Korea (Kim et al.. 2007). An overall TCEP removal of 9 percent was calculated from the removals
Page 484 of 638
-------
reported for two small-, three medium-, and two large-sized municipal sewage treatment plants in
Sweden (Marklund et al.. 2005a). An overall TCEP removal of-19.1 percent was reported from an
activated sludge plant in Albany, New York, based on measured concentrations in wastewater and
suspended particle matter (Kim et al.. 2017). This study was selected for use in this risk evaluation
because this is the best representative of the full-scale wastewater treatment processes that are used in
the United States.
Several high-quality studies observing the efficacy of advanced (tertiary) treatment techniques were
identified. Cristate et al. (2016) reported a low TCEP removal rate (<38%) after a several series of
advanced treatment techniques such as chlorination, ozonation, ultraviolet (UV) radiation, and
UV/hydrogen peroxide (UV/H2O2). Liang and Liu (2016) reported an overall TCEP removal of-30.1
percent after undergoing tertiary treatment that consisted of hyperfiltration, ozonation, and chlorination.
Pang et al. (2016) reported an overall TCEP removal of 0.3 percent and 12.3 percent using UV filters in
two activated sludge plants in China.
Overall, because TCEP has a high water solubility and remains in treated wastewater, negligible to low
accumulation of TCEP will be found in sewage sludge and will not significantly contribute to the
removal of TCEP in wastewater treatments (Kim et al.. 2017; Cristate et al.. 2016; Liang and Liu. 2016;
Marklund et at.. 2005a). In addition, biodegradation and air stripping are not expected to be significant
removal processes. Therefore, TCEP is expected to pass through wastewater treatment systems and be
discharged into the receiving waters.
F.2.5.3 Removal in Drinking Water Treatment
In the United States, drinking water typically comes from surface water (i.e., lakes, rivers, reservoirs)
and groundwater. The source water then flows to a drinking water treatment plant (DWTP) where it
undergoes a series of water treatment steps before being dispersed to homes and communities. In the
United States, public water systems often use "conventional treatment" processes that include
coagulation, flocculation, sedimentation, filtration, and disinfection, as required by law.
Five U.S. studies were identified and reviewed on the removal of TCEP in DWTPs. Those DWTPs
consisted of both conventional and advanced treatment processes and used river water as the source. In
all five studies, TCEP was found to be either minimally removed or not removed at all after undergoing
pre-ozonation (or coagulation), flocculation, sedimentation, ozonation, filtration, and chlorination (Choo
and Oh. 2020; Zhang et al.. 2016a; Benotti et al.. 2009; Snyder et al.. 2006; Westerhoff et al.. 2005;
Stackelberg et al.. 2004).
Several studies have demonstrated that granular activated carbon (GAC) or powdered activated carbon
(PAC) enhanced the removal of TCEP when added to conventional treatment methods (Feng et al..
2023; Choo and Oh. 2020; Padhye et al.. 2014; Westerhoff et al.. 2005; Stackelberg et al.. 2004). A
South Korean drinking water treatment study reported a removal efficiency of 52 percent after
undergoing coagulation and ultrafiltration. After undergoing the GAC step, 73.7 percent of TCEP was
removed (Kim et al.. 2007). Notably, a high level of uncertainty exists about TCEP's carbon usage rate.
The higher the carbon usage rate, the more expensive the treatment costs will be to achieve high levels
of TCEP removal. Higher treatment costs may determine that GAC nor PAC is not an economically
feasible method for removing TCEP from drinking water. In addition, the use of activated carbon
filtration, such as PAC and GAC, is not mandatory for drinking water treatment facilities in the United
States.
Page 485 of 638
-------
F.2.6 Bioaccumulation Potential
Information on bioconcentration and bioaccumulation in aquatic and terrestrial organisms are important
to understand the behavior of TCEP in the environment and a key component in assessing its risk to all
living organisms, including humans.
Bioconcentration is the uptake and retention of a chemical by an aquatic organism from ambient water
only (U.S. EPA. 2003c). Bioconcentration does not include chemical exposure through diet, but rather
its uptake by respiratory and dermal surfaces (Arnot and Gobas. 2006). The bioconcentration factor
(BCF) is the ratio of the concentration of a chemical in the tissue of an organism to its concentration in
the ambient water once a steady state has been achieved (OECD. 2012). The resulting BCF value
provides an indication of the potential for a chemical to bioconcentrate in lipids of organisms.
Three high-quality semi-static tests were identified and selected for use in the risk evaluation. Tang et al.
(2019) reported steady-state BCF values of 1.0 in the muscle, 1.6 in the gill, 2.6 in the brain, 1.6 in the
kidney, and 4.3 in the liver in juvenile common carp (Cyprimis carpio) after 28 days of exposure to
TCEP at 9.1 |ig/L using OECD TG 305 (OECD. 2012). Wang et al. (2017a) reported steady-state BCF
values of 0.8 in the muscle, 1.9 in the gill, 2.2 in the brain, and 2.4 in liver of adult zebrafish (Danio
rerio) after 19 days of exposure to TCEP at 893 |ig/L using OECD TG 305 (OECD. 2012). The
concentration of TCEP in all tissue compartments achieved steady-state in 3 days and the depuration
half-life was less than 5.3 hours. Another high-quality semi-static test reporting BCF values in fish was
identified and selected. Arukwe et al. (2018) reported BCF values of 0.31, 0.16, and 0.34 in the muscle
in juvenile Atlantic salmon (Salmo solar) after 7 days of exposure to TCEP at concentrations of 0.04,
0.2, 1 mg/L, respectively.
A continuous flow-through test was identified during systematic review. Sasaki et al. (1982) reported
BCF values of 1.1 and 1.3 in killifish (Oryzicis latipes) after 5 and 11 days of exposure to TCEP at
concentrations of 12.7 and 2.3 mg/L, respectively. The depuration half-life was 0.7 hour, which
indicates that the killifish eliminated TCEP rapidly. This study was evaluated as a medium-quality study
because insufficient information was available on the test conditions and study design. This added
uncertainty on whether its BCF values would be a good representation of TCEP's bioconcentration
potential and thus will not be considered in this risk evaluation.
The range of experimental BCF values provided above agrees with the calculated BCF values of 1.04
L/kg given by the BCFBAF™ module in EPI Suite™ (U.S. EPA. 2017a) and 1.29 by another QSAR
model, OPEn structure-activity/property Relationship App (OPERA) (U.S. EPA. 2019c; Mansouri et al..
2018). The calculated values from EPI Suite™ and OPERA are not included in this risk evaluation
because the BCF values from high-quality studies cited above are available.
Bioaccumulation is the net accumulation of a chemical by an organism by all possible routes of
exposure (e.g., respiration, dietary, dermal) from all surrounding environmental media (e.g., air, water,
sediment, and diet) (ECHA. 2008). The bioaccumulation factor (BAF) can be expressed as the steady-
state ratio of the chemical concentration in an organism to the concentration in the ambient water. The
concentration of a chemical in an organism can be measured and reported on wet weight, dry weight, or
lipid weight basis. In order to reduce any variability and uncertainty, lipid-normalized BAFs in whole
fish and fish tissues were used in this risk evaluation. Lipid weight BAF (BAFiw) values were converted
to wet weight BAF (BAFww) values by using Equation Apx F-l.
Page 486 of 638
-------
EquationApx F-l.
BAFWW = BAFlw X (^)
There are multiple wet weight BAF values reported for aquatic organisms collected from water bodies
that contained TCEP. A mean BAF value (L/kg wet weight) of 794 in the muscle and 1,995 in the liver,
kidney, and gill, respectively, were reported for pelagic and benthic fish collected from Laizhou Bay in
China (Bekele et al.. 2021). A mean BAF value (L/kg wet weight) of 30.7 in the muscle and 70.7 in the
liver was reported for crucian carp (Carassius auratus) collected from Nakdong River in South Korea
(Choo et al.. 2018). A mean BAF value (L/kg wet weight) of 2,198 was reported in walleye (Sander
vitreus) collected from the Great Lakes (Guo et al.. 2017b). Mean whole body BAF values (L/kg wet
weight) ranging from 109 to 1,248 were reported for aquatic organisms collected from a freshwater pond
containing e-waste in South China (Liu et al.. 2019a). Mean BAF values of 6,310 in benthic
invertebrates, 2,690 in pelagic fish, and 4,270 in benthic fish were reported for fish collected from
Zhushan Bay in Lake Taihu, China (Wang et al.. 2019b).
Zhang et al. (2018b) reported a median BAF value (L/kg wet weight) of 21,380 in the muscle of fishes
collected from a site that was less than 1 km away from the outfall of a wastewater treatment plant
located in Pearl River Delta, China. Fish species included catfish (Clarias batrachus), common carp
(Cyprinus carpi o), bream (Par abramis pekinensis), and white semiknife-carp (Hemiculter leu risen his).
This BAF value is not included in this risk evaluation because this study was evaluated as a medium
quality. Surface water samples were collected from 11 different sites, while fish samples were collected
from only 1 site. Because the TCEP concentrations in surface water were reported as a range,
independent calculation of the BAF could not be conducted. In addition, the reported BAF value could
not be verified whether it was a lipid-normalized BAF value. Hou et al. (2017) reported a mean whole
body BAF value (L/kg wet weight) of 34.7 for topmouth gudgeon, (Pseudorasboraparva), crucian carp
(Carassius auratus), and loach (Misgurnus anguillicaudatus) collected from urban surface water in
Beijing, China. Because this study was evaluated as a medium quality, these BAF data are not included
in this risk evaluation. The tissue-specific values were based on average water concentrations; however,
the study did not specify which of the nine rivers the tissue concentrations in the fish were from and not
all loach samples have reported corresponding concentrations in several rivers, which adds uncertainty
in the study's calculations. Sutton et al. (2019) measured TCEP in the blubber of harbor seals (Phoca
vitulina) from San Francisco Bay. This study is not included in this risk evaluation because upper
trophic fish are the focus of this bioaccumulation assessment.
The upper-trophic fish BAF value of 6.3 and a biotransformation half-life of 0.0798 days (~1 hour and
55 minutes) were estimated using a log Kow value of 1.78 (ECB. 2009) in the BCFBAF™ Model (U.S.
EPA. 2017a). The biotransformation half-life of 0.219 days (-5.3 hours) was estimated by OPERA (U.S.
EPA. 2019c; Mansouri et al.. 2018). These estimated values were not included in this risk evaluation
because data from high-quality monitoring studies are available.
Bioaccumulation from soil to terrestrial or benthic organisms is expressed by the biota-sediment
accumulation factor (BSAF), which is the ratio of concentrations of a chemical in the tissue of a
sediment-dwelling organism to the concentration of a chemical in sediment. Wang et al. (2019b)
reported a BSAF value of 2.19x 10~3 and 1,48x 10~3 for invertebrates and benthic fishes, respectively,
from Zhushan Bay in Lake Taihu, China. Liu et al. (2019a) reported a BSAF range of 0.015 to 0.171 for
aquatic organisms collected from freshwater pond polluted with e-wastes in South China. Choo et al.
(2018) reported a mean BSAF value of 1.09 in the muscle and 2.49 in the liver of crucian carp
(Carassius auratus). Zhang et al. (2018b) reported a BSAF value of 1.38x 10 3 in fish muscles collected
from a site that was less than 1 km away from the outfall of a wastewater treatment plant located in Pearl
Page 487 of 638
-------
River Delta, China. This BSAF value is not included in this risk evaluation because this study was
evaluated as a medium quality. Sediment samples were collected from 11 different sites, while fish
samples were collected from only 1 site. Because the TCEP concentration in sediment was reported as a
range, independent calculation of BSAF could not be conducted.
Biomagnification describes the potential of a chemical to be transferred through the food web. It is
defined as an increase of a chemical concentration in the tissue of an organism compared to the tissue
concentration of its prey. The biomagnification potential of a chemical can be expressed as either a
biomagnification factor (BMF) or trophic magnification factor (TMF). Generally, TMF is preferred over
BMF because TMF represents the average value of the prey-to-predator magnification factor over a food
chain rather than just a specific predator-prey relationship (Fu et al.. 2020). When a trophic dilution
occurs, the concentration of a pollutant decreases as the trophic level increases. It could be a result of a
net balance of ingestion rate, uptake from food, internal transformation, or elimination processes
favoring loss of pollutant that enters the organism via food.
In Brandsma et al. (2015). TMFs were calculated for organophosphate flame retardants (OPFRs) in two
food webs (benthic and pelagic) and in total food web of Western Scheldt in Netherlands. No significant
relationship was observed between TCEP and pelagic food web and total food web. It is possible that the
trophic dilution in the pelagic food web occurred because TCEP was likely to be adsorbed to particles,
and thus were likely to be more abundant in the sediment than in the water column. However, a TMF
value of 2.6 was reported for benthic food web. It was determined that the trophic magnification in the
benthic food web of TCEP was due to high levels of TCEP emission and the organisms' substantial
exposure. Fu et al. (2020) studied the trophic magnification behavior of organophosphate esters in the
Antarctic ecosystem that included algae (Halymenia floresia), archaeogastropoda (Nacella conchma),
neogastropoda (Trophon geversicmus), black rockcod (Notothenia coriiceps), and penguins (Pygoscelis
papua). The TMF of TCEP was 5.2, which indicated that TCEP can be magnified through this food
chain. Zhao et al. (2018) studied the trophic transfer of OPFRs in a lake food web from Taihu Lake,
China, which included plankton, 5 invertebrate species, and 11 fish species. There was no significant
correlation between TCEP and trophic level. Trophic dilution was likely to be a result of rapid
metabolism in sampled fishes.
F.2.6.1 Key Sources of Uncertainty
There is a significant disparity between the BCF and BAF values reported for TCEP. It was observed
that field-measured BAFs were much higher than laboratory-measured BCFs. In controlled laboratory
studies, the exposure time is short, reaching equilibrium is challenging, and the exposure pathway is
limited (lack of dietary intake). A field-measured BAF considers an organism's exposure to a chemical
through all exposure routes in a natural aquatic ecosystem and incorporates chemical biomagnification
and metabolism, making it the most direct measure of bioaccumulation (U.S. EPA. 2003c). TCEP has
the ability to quickly bioaccumulate in fish tissue if it is exposed to high TCEP concentration in the
surrounding water for a period of time. For example, TCEP concentration in the muscle of juvenile
Atlantic salmon (Scilmo solar) increased 10-fold when the water concentration of TCEP increased from
0.2 to 1 mg/L in 7 days (Arukwe et al.. 2018).
Overall, a significantly higher concentration of TCEP was observed in liver than in the muscle (Tang et
al.. 2019; Choo et al.. 2018; Hou et al.. 2017; Wang et al.. 2017a). Hou et al. (2017) showed that
metabolically active tissues, such as liver and kidney, accumulate more than metabolically inactive
tissue like muscle. The liver is the first tissue to be perfused by trace pollutants and it has a higher lipid
contents and assimilation rate than in muscles (Kim et al.. 2015; Koiadinovic et al.. 2007). Several
studies showed that a significant correlation was observed between lipid contents and TCEP
Page 488 of 638
-------
concentrations, indicating that lipid content is an important factor determining TCEP bioaccumulation in
aquatic organisms (Bekele et al.. 2019; Wang et al.. 2017a; Gao et al.. 2014). However, some studies
showed no significant correlations between TCEP concentrations and lipid contents (Liu et al.. 2019a;
Liu et al.. 2019b; Brandsma et al.. 2015). The accumulative potential of TCEP can vary greatly due to
several factors such as fish species, feeding habits, and temporal and spatial factors (U.S. EPA. 2003c).
Collectively, the above studies indicate that TCEP could have the potential to bioaccumulate and
biomagnify in benthic food webs.
The reported TMF reported by Brandsma et al. (2015) was reported as "tentative" because the sample
size was small (n = 15). As a general rule, a number of samples between 30 and 60 are recommended to
achieve statistical reliable TMFs (Borga et al.. 2012). The small sample size adds some uncertainty with
the use of this TMF value in this risk evaluation.
Page 489 of 638
-------
Appendix G ENVIRONMENTAL HAZARD DETAILS
G.l Approach and Methodology
For aquatic species, EPA estimates hazard by calculating a COC for a hazard threshold. COCs can be
calculated using a deterministic method by dividing a hazard value by an AF according to EPA methods
as shown below in Equation_Apx G-l (U.S. EPA 2016a. 2014b. 2012b).
EquationApx G-l.
toxicity value
COCs can also be calculated using probabilistic methods. For example, an SSD can be used to calculate
a hazardous concentration for 5 percent of species (HC05). The HC05 estimates the concentration of a
chemical that is expected to protect 95 percent of aquatic species. This HC05 can then be used to
calculate a COC. For TCEP, Web-ICE, Version 3.3 (Appendix G.2.1.1) followed by SSD probabilistic
method (Appendix G.2.1.4) was used to calculate the acute COC. The deterministic method was used to
calculate at chronic COC
For terrestrial species, EPA estimates hazard by using a hazard value for soil invertebrates, a
deterministic approach, or by calculating a TRV for mammals (Appendix G.2.20). The TRV is
expressed as doses in units of mg/kg-bw/day. Although the TRV for TCEP is derived from laboratory
mice and rat studies, body weight is normalized; therefore, the TRV can be used with ecologically
relevant wildlife species to evaluate chronic dietary exposure to TCEP (U.S. EPA. 2007a).
C.2 Hazard Identification
G.2.1 Aquatic Hazard Data
G.2.1.1 Web-Based Interspecies Correlation Estimation (Web-ICE)
Results from the systematic review process indicated studies with empirical data meeting evaluation
criteria on aquatic species for TCEP with several studies producing LC50 and EC50 endpoint data. To
supplement the empirical data, EPA used a modeling approach, Web-ICE. Web-ICE predicts toxicity
values for environmental species that are absent from a dataset and can provide a more robust dataset to
estimate toxicity thresholds. Specifically, EPA used Web-ICE to supplement empirical data for aquatic
organisms for acute exposure durations for invertebrates and vertebrates. The Agency also used Web-
ICE to supplement empirical hazard data within aquatic plants. EPA also considered ECOSAR
predictions. However, after comparing predictions with empirical data available for TCEP, EPA had
greater confidence in the Web-ICE predictions. Therefore, Web-ICE predictions were used
quantitatively during evidence integration.
G.2.1.2 Invertebrate and Vertebrate Web-ICE
Acute dose-response assays for fish and aquatic invertebrates create useful hazard endpoints for risk
assessments. Calculated endpoints such as EC50 or LC50 values and associated descriptors (CI, NOEC,
and LOEC values) are often comparable across taxa when standardized methodologies and statistical
analysis are employed and documented. Three studies in the TCEP dataset had LC50 data for rainbow
trout, zebrafish, and Daphnia magna (Alzualde et al.. 2018; Torav Research Center. 1997a; Life
Sciences Research Ltd. 1990a) and could be used as "surrogate species" for Web-ICE predictions. This
dataset for aquatic organisms contained data gaps that EPA looked to fill using other lines of evidence
Page 490 of 638
-------
{i.e., modeling approaches). The remaining empirical LC50 values reported in Zhang et al. (2024)
represented four species (brine shrimp, Japanese seabass, Manila clam, mysid shrimp [.Neomysis
awatschensis]) that were currently not available for surrogate to predicted species but were incorporated
into the SSD.
The Web-ICE application was developed by EPA and collaborators to provide interspecies extrapolation
models for acute toxicity (Raimondo and Barron. 2010). Web-ICE models estimate the acute toxicity
(LC50/LD50) of a chemical to a species, genus, or family with no test data (the predicted taxon) from
the known toxicity of the chemical to a species with test data (the commonly tested surrogate species).
Web-ICE models are log-linear least square regressions of the relationship between surrogate and
predicted taxon based on a database of acute toxicity values; that is, median effect or lethal water
concentrations for aquatic species (EC50/LC50). Separate acute toxicity databases are maintained for
aquatic animals (vertebrates and invertebrates), aquatic plants (algae), and wildlife (birds and
mammals), with 1,440 models for aquatic taxa and 852 models for wildlife taxa currently included in
Web-ICE version 3.3 (Willming et al.. 2016). Open-ended toxicity values (i.e., >100 mg/kg or <100
mg/kg) and duplicate records among multiple sources are not included in any of the databases.
The aquatic animal database within Web-ICE comprises 48- or 96-hour EC50/LC50 values based on
death or immobility. This database is described in detail in the Aquatic Database Documentation found
on the Download Model Data page of Web-ICE and describes the data sources, normalization, and
quality and standardization criteria (e.g., data filters) for data used in the models. Data used in model
development adhered to standard acute toxicity test condition requirements of the ASTM International
(ASTM. 2014) and EPA's OCSPP (e.g., (U.S. EPA 2016a)).
EPA used the 96-hour LC50 toxicity data from rainbow trout and zebrafish studies and 48-hour EC50
toxicity data from Daphnia magna in Table 4-2 as surrogate species to predict acute toxicity values
using the Web-ICE application (Raimondo et al.. 2023; Raimondo and Barron. 2010). The Web-ICE
Model estimated toxicity values for 97 species. For model validation, the predicted species model results
are then screened by the following criteria detailed within Willming et al. (2016) to ensure confidence in
the model predictions. If a predicted species did not meet all the quality criteria listed below, the
predicted hazard value from that was not included within the dataset for the SSD:
• High R2 (> -0.6)
o The proportion of the data variance that is explained by the model. The closer the R2
value is to 1.0, the more robust the model is in describing the relationship between the
predicted and surrogate taxa.
• Low mean square error (MSE; < ~0. 95)
o An unbiased estimator of the variance of the regression line.
• High slope (> -0.6)
o The regression coefficient represents the change in log 10 value of the predicted taxon
toxicity for every change in loglO value of the surrogate species toxicity.
• Narrow 95 percent CIs
o Two orders of magnitude between lower and upper limit
After screening, the predicted acute toxicity values for 46 additional aquatic organisms (22 fish, 1
amphibian, 9 aquatic invertebrates, 14 benthic invertebrate species) were added from the surrogate
rainbow trout, zebrafish, and Daphnia magna data (Table Apx G-l). The toxicity data were then used to
Page 491 of 638
-------
calculate the distribution of species sensitivity to TCEP exposure through the SSD toolbox as shown in
FigureApx G-4 and Table 4-6 (Etterson. 2020). The distribution of acute hazard values from
vertebrates and invertebrates was examined to determine if these two groups had similar sensitivity to
TCEP. The mean LC50 values for the 26 vertebrates and 27 invertebrates were 399 ± 54 mg/L (± SEM)
and 197 ±51 mg/L and were not significantly different (F = 1.87, df = 1,51, P = 0.17). Two predicted
values, representing one vertebrate (cape fear shiner) and one invertebrate (tubifex worm), were each
four standard deviations from their respective means. When statistical analysis was conducted with these
two values omitted as outliers the two groups were not significantly different (F = 6.24, df = 1,49, P =
0.015) with mean LC50 values of 253 ± 30 mg/L and 152 ± 27 mg/L for vertebrates and invertebrates,
respectively. Acute hazard values for invertebrates divided between benthic and water column species
were not significantly different (F = 1.01, df = 1,25,P = 0.32) with mean LC50 values of 243 ± 90 mg/L
and 138 ± 27 mg/L, respectively.
G.2.1.3 Algal Web-ICE
Two studies in the TCEP dataset contained EC50 data for two marine diatom species: Phaeodactylum
tricornutum and Skeletonema costatam (Zhang et al.. 2024) as well as a freshwater green alga species
(Raphidocelis subcapitata) (Torav Research Center. 1997b) that could be applied as "surrogate species"
for Web-ICE predictions.
This dataset for aquatic plants contained data gaps that EPA looked to fill using other lines of evidence
(i.e., modeling approaches). The remaining empirical EC50 values reported in Zhang et al. (2024)
represented two marine green algae species (Danaliella salina and Platymonas sabcordiformis)
currently not available for surrogate to predicted species within Web-ICE but were incorporated into the
SSD.
The Web-ICE Model estimated toxicity values for 10 species. For model validation, the predicted
species model results are then screened by the following criteria detailed within Willming et al. (2016)
to ensure confidence in the model predictions. If a predicted species did not meet all the quality criteria
previously listed in Appendix G.2.1.2, the predicted hazard value from that was not included within the
dataset for the SSD. After screening, the predicted acute toxicity values for three additional species
(Desmodesmus subspicatus, Minutocellus polymorphus, and Thalcissiosira pseadonana) were added
from the surrogate Phaeodactylum tricornatam, Skeletonema costatam, and Raphidocelis subcapitata
data (Table Apx G-l). The toxicity data were then used to calculate the distribution of species
sensitivity to TCEP exposure through the SSD toolbox as shown in Figure Apx G-4 and Table 4-6
(Etterson. 2020).
Page 492 of 638
-------
Table Apx G-l. Invertebrate and Vertebrate Web-ICE Predicted Species that Met Model Selection Criteria
Predicted Species
Surrogate Species
LC50
(mg/L)
95% CI
R2
MSE
Slope
Daphnia magna
171.0
Rainbow trout
249.0
Zebrafish embryo
279.1
Amphipod (Grammarus fasciatus)
Daphnia magna
105.70
30.27-369.10
0.75
0.77
0.86
Atlantic salmon
Rainbow trout
260.08
104.18-649.30
0.95
0.12
1.01
Beaver-tail fairy shrimp
Daphnia magna
102.10
64.38-161.91
0.98
0.05
0.91
Bluegill11
Rainbow trout
231.66
183.95-291.73
0.88
0.21
0.93
Bluegill11
Daphnia magna
68.99
45.26-105.17
0.62
0.80
0.66
Brook trout
Rainbow trout
258.83
127.67-524.75
0.94
0.11
1.02
Brown trout
Rainbow trout
252.59
117.39-543.50
0.95
0.10
0.99
Bullfrog11
Rainbow trout
333.43
159.02-699.15
0.97
0.15
0.88
Bullfrog11
Daphnia magna
195.05
34.49-1,102.98
0.86
0.9
0.99
Cape Fear shiner
Rainbow trout
1,443.70
318.31-6,547.86
0.99
0.01
1.18
Channel catfish
Rainbow trout
172.56
100.50-296.29
0.79
0.40
0.82
Channel catfish
Zebrafish embryo
151.86
22.10-1,043.50
0.92
0.09
0.76
Chinook salmon
Rainbow trout
229.96
123.71-427.43
0.96
0.07
0.94
Coho salmon
Rainbow trout
319.44
220.60-462.55
0.98
0.04
0.98
Common carp
Rainbow trout
304.89
104.49-889.56
0.87
0.30
0.89
Cutthroat trout
Rainbow trout
168.04
99.52-283.73
0.94
0.09
0.93
Daphnid (Ceriodaphnia dubia)
Daphnia magna
121.66
70.17-210.92
0.95
0.26
1.00
Daphnid {Daphniapulex)
Daphnia magna
154.64
72.16-331.36
0.97
0.12
1.01
Daphnid (Simocephahts serrulatus)
Daphnia magna
176.58
15.97-1,951.88
0.88
0.21
1.00
Page 493 of 638
-------
Predicted Species
Surrogate Species
LC50
(mg/L)
95% CI
R2
MSE
Slope
Daphnid (Simocephahts vetulus)
Daphnia magna
337.13
298.97-380.16
0.99
0.00
0.98
Fathead minnow11
Rainbow trout
298.23
192.71-461.53
0.83
0.31
0.86
Fathead minnow11
Zebrafish embryo
258.44
135.54-492.77
0.84
0.54
0.91
Fatmucket
Daphnia magna
57.968
20.38-164.81
0.86
0.47
0.74
Fountain darter
Rainbow trout
514.72
37.20-7,120.23
0.99
0
1.09
Goldfish
Rainbow trout
392.65
153.71-1,002.99
0.86
0.42
0.85
Green sunfish
Rainbow trout
314.51
107.19-922.85
0.94
0.13
0.92
Guppy
Rainbow trout
195.83
46.64-822.14
0.73
0.54
0.80
Isopod (Caecidotea brevicauda)
Rainbow trout
33.26
3.60-306.95
0.64
0.59
0.76
Lake trout
Rainbow trout
98.62
51.81-187.73
0.93
0.08
0.86
Largemouth bass
Rainbow trout
143.43
52.46-392.13
0.86
0.24
0.94
Medaka
Rainbow trout
691.06
49.03-9,739.30
0.92
0.32
1.02
Midge (I'aratanytarsus dissimilis)
Rainbow trout
607.00
92.95-3,963.87
0.83
0.54
0.80
Midge (Paratanytarsusparthenogeneticus)
Daphnia magna
449.18
251.62-801.83
0.98
0.04
0.93
Mysid (Americamysis bahia)
Daphnia magna
25.01
12.98-48.18
0.68
0.93
0.83
Mysid (Metamysidopsis insidaris)
Daphnia magna
274.59
20.01-3,777.57
0.94
0.18
0.86
Neosho mucket
Daphnia magna
83.59
6.28-1,112.01
0.97
0.07
0.97
Oligochaete (Tubifex tabifex)
Daphnia magna
1,356.76
196.86-9,350.73
0.87
0.5
0.86
Paper pondshell
Daphnia magna
76.370
42.62-136.83
0.96
0.11
0.90
Sheepshead minnow
Rainbow trout
101.20
47.13-217.30
0.65
0.56
0.75
Shortnose sturgeon
Rainbow trout
474.09
104.63-2,148.02
0.98
0.04
1.14
Swamp lymnaea
Daphnia magna
194.54
81.47-464.51
0.96
0.19
1.01
Tadpole physa
Daphnia magna
146.40
68.29-313.81
0.96
0.14
0.99
Page 494 of 638
-------
Predicted Species
Surrogate Species
LC50
(mg/L)
95% CI
R2
MSE
Slope
Threeridge
Daphnia magna
31.91
13.87-73.42
0.94
0.18
0.87
Vernal pool fairy shrimp
Daphnia magna
105.95
45.16-248.57
0.98
0.09
0.9
Walleye
Rainbow trout
20.30
1.24-332.56
0.67
0.27
0.69
Washboard
Daphnia magna
64.97
32.71-129.02
0.96
0.16
0.92
Water flea (Pseudosia ramosa)
Daphnia magna
43.45
3.94-478.20
0.87
0.57
0.93
Western pearl shell
Daphnia magna
82.23
29.32-171.98
0.95
0.14
0.86
White heel splitter
Daphnia magna
55.66
22.90-135.31
0.98
0.10
0.92
Yellow perch
Rainbow trout
201.79
78.70-517.39
0.94
0.14
0.98
11 The geometric mean of LC50 data for multiple predictions from different surrogate species are used for the species sensitivity distribution (SSD).
Table Apx G-2. Algal Web-ICE Predicted Species that Met Model Selection Criteria
Predicted Species
Surrogate Species
LC50 (mg/L)
95% CI
R2
MSE
Slope
Phaeodactylum tricornutum
76
Skeletonema costatum
353
Rctphidocelis subcapitata
212
Desmodesmus subspicatusa
Phaeodactylum tricornutum
27.6
2.03-374.60
0.93
0.72
0.95
Desmodesmus subspicatusa
Skeletonema costatum
1,153.63
296.57-4,487.48
0.95
0.58
1.01
Desmodesmus subspicatus
Raphidocelis subcapitata
291.25
140.60-603.32
0.96
0.31
1.10
Minutocellus polymorphus
Skeletonema costatum
221.46
23.40-2,095.99
0.75
0.24
1.36
Thalassiosira pseudonana
Skeletonema costatum
411.18
103.73-1,629.93
0.92
0.11
0.7
a The geometric mean of LC50 data for multiple predictions from different surrogate species are used for the species sensitivity distribution (SSD)
Page 495 of 638
-------
G.2.1.4 Species Sensitivity Distribution (SSD)
The SSD Toolbox is a resource created by EPA's Office of Research and Development (ORD) that can
fit SSDs to environmental hazard data (Etterson. 2020). The SSD Toolbox runs on Matlab 2018b (9.5)
for Windows 64 bit. For the TCEP Risk Evaluation, EPA calculated an SSD with the SSD Toolbox
using acute LC50 hazard data from systematic review and estimated data from the Web-ICE application
(Appendix G.2.1.1) that included 25 fish, 1 amphibian, 12 aquatic invertebrates, and 15 benthic
invertebrates. A second SSD was performed for the algae hazard data and was applied with the 5
empirical values and 3 predicted species values detailed in the previous section. The SSD is used to
calculate a HC05. The HC05 estimates the concentration of TCEP that is expected to be protective for
95 percent of species.
G.2.1.5 Vertebrate and Invertebrate SSD
The SSD toolbox contains functions for fitting six distributions (normal, logistic, triangular, Gumbel,
Weibull, and Burr). Maximum likelihood was used to assess the goodness-of-fit of the data distribution
based on Bayesian P-values. The larger the deviation of the p-value from 0.5 the greater the indication
of lack of fit. The Burr distribution P value was nearest to 0.5 at P = 0.51 (Figure Apx G-l), however,
the triangular distribution (P = 0.73) demonstrated the best fit model due to with sample-size corrected
Akaike Information Criterion (AICc) followed by the normal distribution (Figure Apx G-2). Because
numerical methods may lack statistical power for small sample sizes, a visual inspection of the data
were also used to assess goodness-of-fit. For the Q-Q plot, the horizontal axis gives the empirical
quantiles, and the vertical axis gives the predicted quantiles (from the fitted distribution). The Q-Q plot
demonstrates a good model fit with the data points in close proximity to the line across the data
distribution. Q-Q plots were visually used to assess the goodness-of-fit for the distributions (Figure Apx
G-3) with the Burr and logistic distributions demonstrating the best fit. The HC05 was similar among
Burr, triangular, and normal at 31.6 mg/L, 33.4 mg/L, and 34.9 mg/L, respectively; however, the Burr
distribution demonstrated the best fit based on Baysian P-value, and visual inspection of the Q-Q plot.
The SSD plot shows the distribution of species sensitivity to TCEP exposure using Burr distribution
with the calculated HC05 of 31.6 mg/L with a 95 percent CI of 16.7 mg/L to 57.0 mg/L (Figure_Apx
G-4).
G.2.1.6 Algal SSD
The SSD toolbox contains functions for fitting six distributions (normal, logistic, triangular, Gumbel,
Weibull, and Burr). Maximum likelihood was used to assess the goodness-of-fit of the data distribution
based on Bayesian P-values. The larger the deviation of the p-value from 0.5 the greater the indication
of lack of fit. The logistic distribution was nearest to 0.5 at P = 0.53, followed by the Weibull
distribution at 0.46 and normal distribution at 0.39 (Figure Apx G-5). The Weibull and logistic
distributions had the lowest AICc at 103.6 and 105.5, respectively (Figure Apx G-6). Q-Q plots were
visually used to assess the goodness-of-fit for the distributions with both the logistic and normal
demonstrating the best fit (Figure Apx G-7). The HC05 for the logistic and normal distributions were
116.2 and 104.2, respectively, with both distributions resulted in the same lower 95 percent CI of the
HC05 at 66 mg/L (Figure Apx G-8).
Page 496 of 638
-------
® SSD Toolbox - ~ X
File Plot
C:\Users\cgreenQ1\OneDrive - Environmental Protection Agency (EPA)\Documents\modeling\SSD\TCEPacuteDa-
Fit Distribution
Distribution
burr
Fitting method
maximum likelihood
Goodness of Fit:
Iterations:
Scaling parameters
~ Scale to Body Weight
Scaling factor:
Target weight:
1.15
1000
100
Toolbox
Status:
Ready
Results:
Distribution
Method
HC05
p
1
normal
ML
34.9330
0.5894
2
logistic
ML
34.5783
0.4326
3
triangular
ML
33.4063
0.7383
4
gumbel
ML
36.9334
0.0100
5
weibull
ML
16.9913
0.2458
6
burr
ML
31.5853
0.5115
FigureApx G-l. SSD Toolbox Interface Showing HC05s and P-Values for Each
Distribution Using Maximum Likelihood Fitting Method Using TCEP's Acute
Aquatic Hazard Data for Vertebrates and Invertebrates (Etterson, 2020)
Page 497 of 638
-------
•4k ModelSelection
X
Percentile of interest:
Model-averaqed HCp:
Model-averaqed SE of HCp:
CV of HCp:
AlCc Table
5
33.8617
6.9754
0.206
Distribution
AICc
delta AICc
Wt
HCp
SE HCp
1
triangular
685.8545
0
0.4135
33.4063
5.2769
2
normal
686.4240
0.5695
0.3111
34.9330
6.8613
3
logistic
687.3892
1.5347
0.1920
34.5783
7.7071
4
burr
689.3992
3.5447
0.0703
31.5853
10.2772
5
weibull
693.7044
7.8499
0.0082
16.9913
5.7044
6
gumbel
694.7032
8.8487
0.0050
36.9334
5.1043
FigureApx G-2. AICc for the Six Distribution Options in the SSD Toolbox for TCEP's Acute
Aquatic Hazard Data for Vertebrates and Invertebrates (Etterson, 2020)
Page 498 of 638
-------
F#1 Normal Quantile Plot (Q-Q) — ~ X
File
Predicted Quantiles
S Burr Quantile Plot (Q-Q) — ~ X
File
a a ^! ~ id
Predicted Quantiles
(5 Logistic Quantile Plot (Q-Q) — ~ X
File
a d^ as
Predicted Quantiles
[7| Triangular Quantite Plot (Q-Q) — ~ X
File
a a -S ~ is
Predicted Quantiles
Figure Apx G-3. Q-Q Plots of TCEP Acute Aquatic Hazard Data for Vertebrates and
Invertebrates with the (A) Normal, (B) Logistic, (C) Burr, and (D) Triangular Distributions
(Etterson, 2020)
Page 499 of 638
-------
0.9
0.8
0.7
= 0.6
ro
&
2
CL
® 0.5
>
jo
D
I 0.4
O
0.3
0.2
0.1
burr-ML
~ HC05
95% CL HC05
0 —
0.5
Cape Fear shjper ZflffVop/s meki
^ Oligochaeti^f
Medaka On/zias latipes.
Midge Paratanytarsusftissimilis^
Fountain darter Etheostoma jjnticola tr
Shortnose sturgeon Acipenser brevirostrum ^
Midge Paratanytarsus parthenogmyelicus f S
Goldfish Carassius auratus jr f
Daphnid Simocephalus veturns m
Coho salmon Oncorhynchus kisuteh
Green sunfish Lepomis cyane'us
Manila clam Ruditapes philippitflrum
Common carp Cyprinus carpio
Zebrafish embryo Danio/erio
Fathead minnow Pimephales promelas
Mysid Metamysidopsis itfularis .
Bullfrog Lithobates catesbeianus .
Atlantic salmon Salrfi salarJ
Brook trout Salvelinus fontinalisf
Brown trout Safno truttf
Rainbow trout Oncorhynchus my kip • /
Chinook salmon Oncorhynchus tshtfaytscha •
Yellow perch Perca flavescens A f
Guppy Poecilia reticulata fa
Swamp lymnaea Lymnaea stagnalis /• /
Daphnid Simocephalus sem/atus /
Dadhina daphnia magna
Cutthroat trout Oncorhynchus clarkii
Clifnnel catfish Ictalurus punctatus
Daphnid Daphnia pulex
Blijegill Lepomis macrochirus
Tadpole physa Physa gyrina
L/rgemouth bass Micropterus salmoides
' Daphnid Ceriodaphnia dubia
wemalpfo! fairy shrimp Branchinecta lynchi
MAmphipod Gammarus fasciatus
fBeavet/tail fairy shrimp Thamnocephalus platyurus
Sheeoshead minnow Cyprinodon variegatus
Lakemout Salvelinus namaycush
artemia Artemia sp.
f Neoshormucket Lampsilis rafinesqueana
West&n pearlshell Margaritifera falcata
' Paper pond shell Utterbackia imbecillis
flashboaal Megalonaias nervosa
F/fmucket Lampsilis siliquoidea
White heeljpiitter Lasmigona complanata
_f>ese seabass Lateolabrax maculatus
rflea Pseffdosida ramosa
fshid shrimpNeomysis awatschensis
r-rod Caecidotea brevicauda
wee edgefan die ma plicata
^ Wysid AmericamySff bahia
Walleye S^nd^ritrmjs | j
1.5
2 2.5
Toxicity Value (Log 10[LC50]) mg/L
3.5
FigureApx G-4. SSD Distribution for TCEP's Acute Hazard Data for Invertebrates and Vertebrates (Etterson, 2020).
Aquatic invertebrate/vertebrate HC05 of 31.6 mg/L with a 95 percent CI of 16.7 mg/L to 57.0 mg/L
Page 500 of 638
-------
*'¦ SSD Toolbox
File Plot
~
X
C:\Users\cgreen01\OneDrive - Environmental Protection Agency (EPA)\Documents\modeling\SSD\TCEPalgae.xls
Fitting method
maximum likelihood
Goodness of Fit:
Iterations: I 1000
Scaling parameters
~ Scale to Body Weight
Scaling factor:
Target weight:
100
1.15
Distribution
Method
HC05
p
1
normal
ML
104.6280
0.3916
2
logistic
ML
116.2775
0.5335
3
triangular
ML
91.4494
0.0539
4
gumbel
ML
94.4956
0.1009
5
weibull
ML
104.4049
0.4655
6
burr
ML
81.6823
0.3704
FigureApx G-5. SSD Toolbox Interface Showing HC05s and P-Values for
Each Distribution Using Maximum Likelihood Fitting Method Using TCEP's
Acute Aquatic Hazard Data for Algae (Stterson, 2020)
Page 501 of 638
-------
¦4 ModelSelection
Percentile of interest:
Model-averaqed HCp:
Model-averaqed SE of HCp:
CV of HCp:
AlCc Table
101.2841
NaN
NaN
X
Distribution
AICc
delta AICc
Wt
HCp
SE HCp
1
weibull
103.6055
0
0.4149
104.4049
38.6730
2
logistic
105.5254
1.9199
0.1588
116.2775
34.4863
3
triangular
105.5738
1.9683
0.1551
91.4494
16.6018
4
normal
106.0010
2.3955
0.1252
104.6280
28.7994
5
burr
106.0331
2.4277
0.1232
81.6823
NaN
6
gumbel
109.4101
5.8047
0.0228
94.4956
21.5653
Figure Apx G-6. AICc for the Six Distribution Options in the SSD Toolbox for TCEP's Acute
Aquatic Hazard Data for Algae (Etterson. 2020)
Page 502 of 638
-------
® Weibull Quantile Plot (Q-Q) - ~ X
File
I uW~
Predicted Quantiles
[5 Normal Quantile Plot (Q-Q) — ~ X
File
aas ~ b
[71 Logistic Quantile Plot (Q-Q) — ~ X
File
a a £ | ~ 0
5] Burr Quantile Plot (Q-Q) — ~ X
File
a a ^ | ~ 0
o. 0.4
E
0.3 0.4 0.5 0.6 0.7
Predicted Quantiles
cl 0.4
E
0.3 0.4 0.5 0.6 0.7
Predicted Quantiles
FigureApx G-7. Q-Q Plots of TCEP Acute Aquatic Hazard Data for Algae with the (A) Weibull,
(B) Logistic, (C) Normal, and (D) Burr Distributions (Etterson. 2020)
Page 503 of 638
-------
Toxicity Value (Log 10[EC50]) mg/L
FigureApx G-8. SSD Distribution for TCEP's Acute Hazard Data for Algae (Etterson. 2020).
Aquatic Algae HC05 for the logistic and normal distributions were 116.2 and 104.2, respectively, with both distributions resulting in a lower
95% CI of 66 mg/L.
Page 504 of 638
-------
G.2.2 Terrestrial Hazard Data
For calculation of the mammal TRV, an a priori framework for selection of the TRV value based on the
results of the NOAEL and LOAEL data (Figure Apx G-9.). The minimum dataset required to calculate
a TRV consists of three results with NOEL or LOEL values for reproduction, growth, or mortality for at
least two species. If these minimum results are not available, then a TRV is not calculated.
For mammalian species, EPA estimates hazard by calculating a TRV. The TRV is expressed as doses in
units of mg/kg-bw/day. Although the TRV for TCEP is derived from laboratory mice and rat studies,
body weight is normalized; therefore, the TRV can be used with ecologically relevant wildlife species to
evaluate chronic dietary exposure to TCEP. Representative wildlife species chronic hazard threshold
will be evaluated in the trophic transfer assessments using the TRV. The flow chart in Figure Apx G-9.
was used to select the data to calculate the TRV with NOEL and/or LOEL data and described below
(U.S. EPA. 2007a).
Step 1: At least three results and two species tested for reproduction, growth, or mortality general
end points.
For rats, a 2-year NOEL/LOEL (NTP. 1991b). a 16-week NOEL/LOEL for males, and a 16-
week NOEL/LOEL for females for mortality were used (Matthews et al.. 1990).
For mice, a 16-week NOEL/LOEL for reproduction (Matthews et al.. 1990) and an 8-day LOEL
for mortality were used (Hazleton Laboratories. 1983).
Step 2: Are there three or more NOELs in reproduction or growth effect groups?
Because there was only a single reproduction effect result and no growth effect results, then
proceed to step 3.
Step 3: If there is at least one NOEL result for the reproduction or growth effect groups?
The NOEL for reproduction is 175 mg/kg-bw/day
Then the TRV is equal to the lowest reported NOEL for any effect group (reproduction, growth,
or mortality), except in cases where the NOEL is higher than the lowest bounded LOEL.
The lowest bounded LOEL for mortality is 88 mg/kg-bw/day
Then the TRV is equal to the highest bounded NOEL below the lowest bounded LOEL.
The highest NOEL below the lowest NOEL is 44 mg/kg-bw/day.
The TRV for TCEP is 44 mg/kg-bw/day.
Page 505 of 638
-------
NO
Stq> 1;
Are chars at leant 3
tcsricityvahjes 5>r 2
ijKcis RE51. GRO or
MOR?
YES
S
-------
The environmental hazard integration is organized around effects to aquatic and terrestrial organisms as
well as the respective environmental compartments (e.g., pelagic, benthic, soil). Environmental hazard
assessment may be complex based on the considerations of the quantity, relevance, and quality of the
available evidence.
For TCEP, environmental hazard data from toxicology studies identified during systematic review have
used evidence that characterizes apical endpoints; that is, endpoints that could have population-level
effects such as reproduction, growth, and/or mortality. Additionally, mechanistic data that can be linked
to apical endpoints will add to the weight of scientific evidence supporting hazard thresholds. EPA also
considered predictions from Web-ICE and ECOSAR to supplement the empirical data found during
systematic review.
G.2.3.1 Weight of Scientific Evidence
After calculating the hazard thresholds that were carried forward to characterize risk, a narrative
describing the weight of scientific evidence and uncertainties was completed to support EPA's
decisions. The weight of scientific evidence fundamentally means that the evidence is weighed (i.e.,
ranked) and weighted (i.e., a piece or set of evidence or uncertainty may have more importance or
influence in the result than another). Based on the weight of scientific evidence and uncertainties, a
confidence statement was developed that qualitatively ranks (i.e., robust, moderate, slight, or
indeterminate) the confidence in the hazard threshold. The qualitative confidence levels are described
below.
The evidence considerations and criteria detailed within U.S. EPA (2021a) guides the application of
strength-of-evidence judgments for environmental hazard effect within a given evidence stream and
were adapted from Table 7-10 of the 2021 Draft Systematic Review Protocol (U.S. EPA 2021a).
EPA used the strength-of-evidence and uncertainties from (U.S. EPA 2021a) for the hazard assessment
to qualitatively rank the overall confidence using evidence Table 4-8 for environmental hazard.
Confidence levels of robust (+ + +), moderate (+ +), slight (+), or indeterminant are assigned for each
evidence property that corresponds to the evidence considerations (U.S. EPA 2021a). The rank of the
Quality of the Database consideration is based on the systematic review overall quality determination
(High, Medium, or Low) for studies used to calculate the hazard threshold, and whether there are data
gaps in the toxicity dataset. Another consideration in the Quality of the Database is the risk of bias (i.e.,
how representative is the study to ecologically relevant endpoints). Additionally, because of the
importance of the studies used for deriving hazard thresholds, the Quality of the Database consideration
may have greater weight than the other individual considerations. The high, medium, and low systematic
review overall quality determination ranks correspond to the evidence table ranks of robust (+ + +),
moderate (+ +), or slight (+), respectively. The evidence considerations are weighted based on
professional judgment to obtain the overall confidence for each hazard threshold. In other words, the
weights of each evidence property relative to the other properties are dependent on the specifics of the
weight of scientific evidence and uncertainties that are described in the narrative and may or may not be
equal. Therefore, the overall score is not necessarily a mean or defaulted to the lowest score. The
confidence levels and uncertainty type examples are described below.
Confidence Levels
• Robust (+ + +) confidence suggests thorough understanding of the scientific evidence and
uncertainties. The supporting weight of scientific evidence outweighs the uncertainties to the
point where it is unlikely that the uncertainties could have a significant effect on the exposure or
hazard estimate.
Page 507 of 638
-------
• Moderate (+ +) confidence suggests some understanding of the scientific evidence and
uncertainties. The supporting scientific evidence weighed against the uncertainties is reasonably
adequate to characterize exposure or hazard estimates.
• Slight (+) confidence is assigned when the weight of scientific evidence may not be adequate to
characterize the scenario, and when the assessor is making the best scientific assessment possible
in the absence of complete information. There are additional uncertainties that may need to be
considered.
• Indeterminant (N/A) corresponds to entries in evidence tables where information is not available
within a specific evidence consideration.
Types of Uncertainties
The following uncertainties may be relevant to one or more of the weight of scientific evidence
considerations listed above and will be integrated into that property's rank in the evidence table (Table
4-8):
• Scenario Uncertainty: Uncertainty regarding missing or incomplete information needed to fully
define the exposure and dose.
o The sources of scenario uncertainty include descriptive errors, aggregation errors, errors
in professional judgment, and incomplete analysis.
• Parameter Uncertainty: Uncertainty regarding some parameter.
o Sources of parameter uncertainty include measurement errors, sampling errors,
variability, and use of generic or surrogate data.
• Model Uncertainty: Uncertainty regarding gaps in scientific theory required to make predictions
on the basis of causal inferences.
o Modeling assumptions may be simplified representations of reality.
Table Apx G-3 summarizes the weight of scientific evidence and uncertainties, while increasing
transparency on how EPA arrived at the overall confidence level for each exposure hazard threshold.
Symbols are used to provide a visual overview of the confidence in the body of evidence, while de-
emphasizing an individual ranking that may give the impression that ranks are cumulative (e.g., ranks of
different categories may have different weights).
Page 508 of 638
-------
TableApx G-3. Considerations that Inform Evaluations of the Strength of the Evidence within an Evidence Stream Apical
Endpoints, Mechanistic, or Field Studies)
Consideration
Increased Evidence Strength (of the Apical
Endpoints, Mechanistic, or Field Studies
Evidence)
Decreased Evidence Strength (of the Apical Endpoints, Mechanistic,
or Field Studies Evidence)
The evidence considerations and criteria laid out here guide the application of strength-of-evidence judgments for an outcome or environmental hazard effect
within a given evidence stream. Evidence integration or synthesis results that do not warrant an increase or decrease in evidence strength for a given
consideration are considered "neutral" and are not described in this table (and, in general, are captured in the assessment-specific evidence profile tables).
Quality of the database'1
(risk of bias)
• A large evidence base of high- or mediiim-qua\ity
studies increases strength.
• Strength increases if relevant species are represented
in a database.
• An evidence base of mostly /ow-quality studies decreases strength.
• Strength also decreases if the database has data gaps for relevant species
(i.e., a trophic level that is not represented).
• Decisions to increase strength for other considerations in this table should
generally not be made if there are serious concerns for risk of bias; in other
words, all the other considerations in this table are dependent upon the
quality of the database.
Consistency
Similarity of findings for a given outcome (e.g., of a
similar magnitude, direction) across independent
studies or experiments increases strength, particularly
when consistency is observed across species,
lifestage, sex, wildlife populations, and across or
within aquatic and terrestrial exposure pathways.
• Unexplained inconsistency (i.e., conflicting evidence; see U.S. EPA
(2005b) decreases strength.)
• Strength should not be decreased if discrepant findings can be reasonably
explained by study confidence conclusions; variation in population or
species, sex, or lifestage; frequency of exposure (e.g., intermittent or
continuous); exposure levels (low or high); or exposure duration.
Strength (effect
magnitude) and precision
• Evidence of a large magnitude effect (considered
either within or across studies) can increase strength.
• Effects of a concerning rarity or severity can also
increase strength, even if they are of a small
magnitude.
• Precise results from individual studies or across the
set of studies increases strength, noting that biological
significance is prioritized over statistical significance.
• Use of probabilistic model (e.g., Web-ICE, SSD)
may increase strength.
Strength may be decreased if effect sizes that are small in magnitude are
concluded not to be biologically significant, or if there are only a few
studies with imprecise results.
Biological gradient/dose-
response
• Evidence of dose-response increases strength.
• Dose-response may be demonstrated across studies
or within studies and it can be dose- or duration-
dependent.
• Dose response may not be a monotonic dose-
response (monotonicity should not necessarily be
• A lack of dose-response when expected based on biological
understanding and having a wide range of doses/exposures evaluated in the
evidence base can decrease strength.
• In experimental studies, strength may be decreased when effects resolve
under certain experimental conditions (e.g., rapid reversibility after
removal of exposure).
Page 509 of 638
-------
Consideration
Increased Evidence Strength (of the Apical
Endpoints, Mechanistic, or Field Studies
Evidence)
Decreased Evidence Strength (of the Apical Endpoints, Mechanistic,
or Field Studies Evidence)
expected (e.g., different outcomes may be expected at
low vs. high doses due to activation of different
mechanistic pathways or induction of systemic
toxicity at very high doses).
• Decreases in a response after cessation of exposure
(e.g., return to baseline fecundity) also may increase
strength by increasing certainty in a relationship
between exposure and outcome (this particularly
applicable to field studies).
• However, many reversible effects are of high concern. Deciding between
these situations is informed by factors such as the toxicokinetics of the
chemical and the conditions of exposure, see (U.S. EPA. 1998b). cndooint
severity, judgments regarding the potential for delayed or secondary
effects, as well as the exposure context focus of the assessment (e.g.,
addressing intermittent or short-term exposures).
• In rare cases, and typically only in toxicology studies, the magnitude of
effects at a given exposure level might decrease with longer exposures
(e.g., due to tolerance or acclimation).
• Like the discussion of reversibility above, a decision about whether this
decreases evidence strength depends on the exposure context focus of the
assessment and other factors.
• If the data are not adequate to evaluate a dose-response pattern, then
strength is neither increased nor decreased.
Biological relevance
Effects observed in different populations or
representative species suggesting that the effect is
likely relevant to the population or representative
species of interest (e.g., correspondence among the
taxa, lifestages, and processes measured or observed
and the assessment endpoint).
An effect observed only in a specific population or species without a clear
analogy to the population or representative species of interest decreases
strength.
Physical/chemical
relevance
Correspondence between the substance tested and the
substance constituting the stressor of concern.
The substance tested is an analogue of the chemical of interest or a mixture
of chemicals which include other chemicals besides the chemical of
interest.
Environmental relevance
Correspondence between test conditions and
conditions in the region of concern.
The test is conducted using conditions that would not occur in the
environment.
" Database refers to the entire dataset of studies integrated in the environmental hazard assessment and used to inform the strength of the evidence. In this context,
database does not refer to a computer database that stores aggregations of data records such as the ECOTOX Knowledgebase.
Page 510 of 638
-------
Appendix H ENVIRONMENTAL RISK DETAILS
H.l Risk Estimation for Aquatic Organisms
TableApx H-l. Calculated RQs Based on TCEP Surface Water Concentrations (ppb) as
Calculated Using Mode
ed Data for Annual Air Deposition to Surface Water
Exposure Scenario
Production
Volume
(lb/year)fl
Meteorological
Model6
Surface Water
Concentration (ppb)
at 1,000 mc
Chronic RQ (Hazard
Value: 2.8 ppb)
2,500
MetCT
3.93E-05
1.40E-05
Import and repackaging
MetHIGH
4.78E-05
1.71E-05
25,000
MetCT
1.40E-04
5.00E-05
MetHIGH
1.94E-04
6.93E-05
Incorporation into paints
and coatings - 1-part
coatings
2,500
MetCT
8.60E-04
3.07E-04
MetHIGH
1.37E-03
4.89E-04
25,000
MetCT
1.95E-03
6.96E-04
MetHIGH
2.07E-03
7.39E-04
Incorporation into paints
and coatings - 2-part
reactive coatings
2,500
MetCT
2.20E-04
7.86E-05
MetHIGH
3.18E-04
1.14E-04
25,000
MetCT
6.05E-04
2.16E-04
MetHIGH
9.65E-04
3.45E-04
Use in paints and
coatings - spray
application
2,500
MetCT
3.42E-01
1.22E-01
MetHIGH
4.93E-01
1.76E-01
25,000
MetCT
5.10
1.82
MetHIGH
8.10
2.89
Formulation of TCEP-
2,500
MetCT
1.02E-03
3.64E-04
containing reactive resins
MetHIGH
9.70E-04
3.46E-04
(for use in 2-part
25,000
MetCT
7.60E-04
2.71E-04
systems)
MetHIGH
7.05E-04
2.52E-04
2,500
MetCT
2.46E-04
8.79E-05
Processing into 2-part
MetHIGH
3.55E-04
1.27E-04
resin article
25,000
MetCT
7.20E-04
2.57E-04
MetHIGH
1.15E-03
4.11E-04
2,500
MetCT
1.26E-03
4.50E-04
Laboratory chemicals
MetHIGH
1.17E-03
4.18E-04
25,000
MetCT
7.20E-04
2.57E-04
MetHIGH
6.65E-04
2.38E-04
11 Production volume of 2,500 lb TCEP/year uses high-end estimates (95th percentile). Production volume of 25,000
lb TCEP/yr uses central tendency estimates (median).
h The ambient air modeled concentrations and deposition values are presented for two meteorology conditions (Sioux
Falls, South Dakota, for central tendency meteorology; and Lake Charles, Louisiana, for higher-end meteorology).
c Estimated annual concentrations of TCEP (90th percentile) that could be in surface water via air deposition at a
community (1,000 m from the source) exposure scenario.
Page 511 of 638
-------
TableApx H-2. Environmental RQs by Exposure Scenario with Production Volumes of 2,500 lb/year with Central Tendency Release
Estimates for Aquatic Organisms with TCEP Surface Water Concentration (ppb) Modeled by VVWM-PSC with 50% Percentile
Flow of the 7Q10
Exposure
Scenario
Production
Volume (lb/year)"
Days of
Release
Release
(kg/day)
Modeled Using WWM-PSCC
Max Day Ave
(ppb)*
COC Type
COC (ppb)
Days of Exceedance
(days per year)
RQ
Import and
repackaging
2,500
4
6.35
1,510
Acute
16,700
N/A
0.09
202
Chronic
2.8
33
72.14
Incorporation into
paints and
coatings - 1-part
coatings
2,500
6
10.23
2,960
Acute
16,700
N/A
0.18
596
Chronic
2.8
74
212.86
Incorporation into
paints and
coatings - 2-part
reactive coatings
2,500
1
27.12
6,930
Acute
16,700
N/A
0.41
264
Chronic
2.8
32
94.28
Use in paints and
coatings at - spray
application
2,500
1
2.36
520
Acute
16,700
NA
0.03
18.8
Chronic
2.8
29
6.71
Formulation of
TCEP into 2-part
reactive resins
2,500
1
25.19
7,220
Acute
16,700
N/A
0.43
290
Chronic
2.8
39
103.57
Laboratory
chemicals
2,500
193
0.88
212
Acute
16,700
N/A
1.27E-02
212
Chronic
2.8
226
75.71
" Production volume of 2,500 lb TCEP/year uses central tendency estimates (50th percentile for all COUs except the laboratory chemicals COU uses the 5th percentile).
b Max day average represents the maximum concentration over a 1 - or 30-day average period corresponding with the acute or chronic COC used for the RQ estimate.
"Flow inputs for PSC represent the 50th percentile 7Q10 flows as the lowest expected weekly flow over a 10-year period.
N/A = Days of exceedance are modeled for the application of chronic COCs and do not apply for acute COCs and corresponding RQs.
Page 512 of 638
-------
TableApx H-3. Environmental RQs by Exposure Scenario with Production Volumes of 2,500 lb/year with Central Tendency Release
Estimates for Aquatic Organisms with TCEP Pore Water Concentration (ppb) Modeled by VVWM-PSCwith 50% Percentile Flow of
the 7Q10
Exposure Scenario
Production
Volume
(lb/year)fl
Days of
Release
Release
(kg/day)
Benthic Pore
Water
Concentration
(ppb)6
Benthic Pore Water Concentration'
COC Type
COC
(PPb)
Days of
Exceedance
RQ
Import and repackaging
2,500
4
6.35
98.6
Acute
16,700
N/A
5.90E-03
79.6
Chronic
2.8
225
28.43
Incorporation into paints and coatings -
1-part coatings
2,500
6
10.23
285
Acute
16,700
N/A
1.70E-02
232
Chronic
2.8
291
82.85
Incorporation into paints and coatings -
2-part reactive coatings
2,500
1
27.12
130
Acute
16,700
N/A
7.78E-03
105
Chronic
2.8
242
37.5
Use in paints and coatings - spray
application
2,500
1
2.36
9.29
Acute
16,700
N/A
5.60E-04
7.48
Chronic
2.8
78
2.67
Formulation of TCEP into 2-part
reactive resins
2,500
1
25.19
141
Acute
16,700
N/A
8.44E-03
115
Chronic
2.8
249
41.07
Laboratory chemicals
2,500
193
0.88
200
Acute
16,700
N/A
1.19E-02
199
Chronic
2.8
364
71.07
" Production volume of 2,500 lb TCEP/year uses central tendency estimates (50th percentile for all COUs except the laboratory chemicals COU uses the 5th
percentile).
b Max day average represents the maximum concentration over a 1- or 30-day average period corresponding with the acute or chronic COC used for the RQ estimate.
"Flow inputs for PSC represent the 50th percentile 7Q10 flows as the lowest expected weekly flow over a 10-year period.
N/A = Days of exceedance are modeled for the application of chronic COCs and do not apply for acute COCs and corresponding RQs.
Page 513 of 638
-------
TableApx H-4. Environmental RQs by Exposure Scenario with Production Volumes of 2,500 lb/year with Central Tendency Release
Estimates for Aquatic Organisms with TCEP Surface Water Concentration (ppb) Modeled by VVWM-PSC with 90% Percentile
Flow of the 7Q10
Exposure
Scenario
Production
Volume
(lb/year)fl
Days of
Release
Release
(kg/day)
Modeled Using WWM-PSCC
Max Day Ave
(ppb)6
COC Type
COC (ppb)
Days of Exceedance
(days per year)
RQ
Import and
repackaging
2,500
4
6.35
8.70
Acute
16,700
N/A
5.20E-04
1.17
Chronic
2.8
0
0.41
Incorporation into
paints and
coatings - 1-part
coatings
2,500
6
10.23
10.60
Acute
16,700
N/A
6.30E-04
2.12
Chronic
2.8
0
0.75
Incorporation into
paints and
coatings - 2-part
reactive coatings
2,500
1
27.12
28.20
Acute
16,700
N/A
1.69E-03
0.94
Chronic
2.8
0
0.33
Use in paints and
coatings - spray
application
2,500
1
2.36
3.30
Acute
16,700
NA
2.00E-04
0.11
Chronic
2.8
0
0.04
Formulation of
TCEP into 2-part
reactive resins
2,500
1
25.19
4.00
Acute
16,700
N/A
2.40E-04
0.13
Chronic
2.8
0
0.05
Laboratory
chemicals
2,500
193
0.88
1.20
Acute
16,700
N/A
7.00E-05
1.21
Chronic
2.8
0
0.43
" Production volume of 2,500 lb TCEP/year uses central tendency estimates (50th percentile for all COUs except the laboratory chemicals COU uses the 5th percentile).
b Max day average represents the maximum concentration over a 1 - or 30-day average period corresponding with the acute or chronic COC used for the RQ estimate.
"Flow inputs for PSC represent the 90th percentile 7Q10 flows as the lowest expected weekly flow over a 10-year period.
N/A = Days of exceedance are modeled for the application of chronic COCs and do not apply for acute COCs and corresponding RQs.
Page 514 of 638
-------
TableApx H-5. Environmental RQs by Exposure Scenario with Production Volumes of 2,500 lb/year with Central Tendency Release
Estimates for Aquatic Organisms with TCEP Pore Water Concentration (ppb) Modeled by VVWM-PSCwith 90% Percentile Flow of
the 7Q10
Exposure Scenario
Production
Volume
(lb/year)fl
Days of
Release
Release
(kg/day)
Benthic Pore
Water
Concentration
(ppb)6
Benthic Pore Water Concentration'
COC Type
COC
(PPb)
Days of
Exceedance
RQ
Import and repackaging
2,500
4
6.35
0.58
Acute
16,700
N/A
3.47E-05
0.46
Chronic
2.8
0
0.16
Incorporation into paints and coatings -
1-part coatings
2,500
6
10.23
1.04
Acute
16,700
N/A
6.23E-05
0.83
Chronic
2.8
0
0.29
Incorporation into paints and coatings -
2-part reactive coatings
2,500
1
27.12
0.48
Acute
16,700
N/A
2.87E-05
0.38
Chronic
2.8
0
0.14
Use in paints and coatings - spray
application
2,500
1
2.36
0.06
Acute
16,700
N/A
3.59E-06
0.04
Chronic
2.8
0
1.42E-02
Formulation of TCEP into 2-part
reactive resins
2,500
1
25.19
0.07
Acute
16,700
N/A
4.19E-06
0.05
Chronic
2.8
0
1.78E-02
Laboratory chemicals
2,500
193
0.88
1.15
Acute
16,700
N/A
6.89E-05
1.14
Chronic
2.8
84
0.41
" Production volume of 2,500 lb TCEP/year uses central tendency estimates (50th percentile for all COUs except the laboratory chemicals COU uses the 5th
percentile).
b Max day average represents the maximum concentration over a 1- or 30-day average period corresponding with the acute or chronic COC used for the RQ estimate.
"Flow inputs for PSC represent the 90th percentile 7Q10 flows as the lowest expected weekly flow over a 10-year period.
N/A = Days of exceedance are modeled for the application of chronic COCs and do not apply for acute COCs and corresponding RQs.
Page 515 of 638
-------
TableApx H-6. Environmental RQs by Exposure Scenario with Production Volumes of 2,500 lb/year with High-End Release
Estimates for Aquatic Organisms with TCEP Surface Water Concentration (ppb) Modeled by VVWM-PSC with 90% Percentile
Flow of the 7Q10
Exposure
Scenario
Production
Volume
(lb/year)fl
Days of
Release
Release
(kg/day)
Modeled Using WWM-PSCC
Max Day Ave
(ppb)6
COC Type
COC (ppb)
Days of Exceedance
(days per year)
RQ
Import and
repackaging
2,500
4
9.88
13.60
Acute
16,700
N/A
8.10E-04
1.81
Chronic
2.8
0
0.64
Incorporation into
paints and
coatings - 1-part
coatings
2,500
2
35.18
36.50
Acute
16,700
N/A
2.19E-03
2.43
Chronic
2.8
0
0.87
Incorporation into
paints and
coatings - 2-part
reactive coatings
2,500
1
31.89
33.10
Acute
16,700
N/A
1.98E-03
1.10
Chronic
2.8
0
0.39
Use in paints and
coatings - spray
application
2,500
2
23.26
32.00
Acute
16,700
NA
1.92E-03
2.13
Chronic
2.8
0
0.76
Formulation of
TCEP into 2-part
reactive resins
2,500
1
31.54
5.00
Acute
16,700
N/A
3.00E-04
0.17
Chronic
2.8
0
0.06
Laboratory
chemicals
2,500
182
0.40
0.60
Acute
16,700
N/A
4.00E-05
0.55
Chronic
2.8
0
0.2
" Production volume of 2,500 lb TCEP/year uses high-end estimates (95th percentile for all COUs except the laboratory chemicals COU uses the 1st percentile).
b Max day average represents the maximum concentration over a 1 - or 30-day average period corresponding with the acute or chronic COC used for the RQ estimate.
"Flow inputs for PSC represent the 90th percentile 7Q10 flows as the lowest expected weekly flow over a 10-year period.
N/A = Days of exceedance are modeled for the application of chronic COCs and do not apply for acute COCs and corresponding RQs.
Page 516 of 638
-------
TableApx H-7. Environmental RQs by Exposure Scenario with Production Volumes of 2,500 lb/year with High-End Release
Estimates for Aquatic Organisms with TCEP Pore Water Concentration (ppb) Modeled by VVWM-PSCwith 90% Percentile Flow of
the 7Q10
Exposure Scenario
Production
Volume
(lb/year)fl
Days of
Release
Release
(kg/day)
Benthic Pore
Water
Concentration
(ppb)6
Benthic Pore Water Concentration'
COC Type
COC
(PPb)
Days of
Exceedance
RQ
Import and repackaging
2,500
4
9.88
0.9
Acute
16,700
N/A
5.39E-05
0.72
Chronic
2.8
0
0.25
Incorporation into paints and coatings -
1-part coatings
2,500
2
35.18
1.23
Acute
16,700
N/A
7.37E-05
0.97
Chronic
2.8
0
0.34
Incorporation into paints and coatings -
2-part reactive coatings
2,500
1
31.89
0.56
Acute
16,700
N/A
3.35E-05
0.44
Chronic
2.8
0
0.16
Use in paints and coatings - spray
application
2,500
2
23.26
1.08
Acute
16,700
N/A
6.47E-05
0.85
Chronic
2.8
0
0.3
Formulation of TCEP into 2-part
reactive resins
2,500
1
31.54
0.08
Acute
16,700
N/A
4.79E-06
0.7
Chronic
2.8
0
0.25
Laboratory chemicals
2,500
182
0.40
0.52
Acute
16,700
N/A
3.11E-05
0.52
Chronic
2.8
0
0.19
" Production volume of 2,500 lb TCEP/year uses high-end estimates (95th percentile for all COUs except the laboratory chemicals COU uses the 5th percentile).
b Max day average represents the maximum concentration over a 1- or 30-day average period corresponding with the acute or chronic COC used for the RQ estimate.
"Flow inputs for PSC represent the 90th percentile 7Q10 flows as the lowest expected weekly flow over a 10-year period.
N/A = Days of exceedance are modeled for the application of chronic COCs and do not apply for acute COCs and corresponding RQs.
Page 517 of 638
-------
TableApx H-8. Environmental RQs by Exposure Scenario with Production Volumes of 25,000 lb/year with Central Tendency
Release Estimates for Aquatic Organisms with TCEP Surface Water Concentration (ppb) Modeled by VVWM-PSC with 50%
Percentile Flow of the 7Q10
Exposure
Scenario
Production
Volume
(lb/year)fl
Days of
Release
Release
(kg/day)
Modeled Using WWM-PSCC
Max Day Ave
(ppb)6
COC Type
COC (ppb)
Days of Exceedance
(days per year)
RQ
Import and
repackaging
25,000
39
7.13
1,710
Acute
16,700
N/A
0.10
1700
Chronic
2.8
156
607.14
Incorporation into
paints and
coatings - 1-part
coatings
25,000
57
10.98
3,220
Acute
16,700
N/A
0.19
3,210
Chronic
2.8
237
1146.43
Incorporation into
paints and
coatings - 2-part
reactive coatings
25,000
4
65.90
19,100
Acute
16,700
N/A
1.14
2560
Chronic
2.8
158
914.29
Use in paints and
coatings - spray
application
25,000
1
2.31
509
Acute
16,700
N/A
0.03
18.4
Chronic
2.8
29
6.57
Formulation of
TCEP into 2-part
reactive resins
25,000
3
45.51
15,500
Acute
16,700
N/A
0.93
1570
Chronic
2.8
139
560.71
" Production volume of 25,000 lb TCEP/year uses central tendency estimates (50th percentile for all COUs).
h Max day average represents the maximum concentration over a 1- or 30-day average period corresponding with the acute or chronic COC used for the RQ estimate.
"Flow inputs for PSC represent the 50th percentile 7Q10 flows as the lowest expected weekly flow over a 10-year period.
N/A = Days of exceedance are modeled for the application of chronic COCs and do not apply for acute COCs and corresponding RQs.
Page 518 of 638
-------
TableApx H-9. Environmental RQs by Exposure Scenario with Production Volumes of 25,000 lb/year with central tendency release
estimates for Aquatic Organisms with TCEP Benthic Pore Water Concentration (ppb) Modeled by VVWM-PSC with 50% percentile
flow of the 7Q10
Exposure Scenario
Production
Volume
(lb/year)fl
Days of
Release
Release
(kg/day)
Benthic Pore
Water
Concentration
(ppb)6
Benthic Pore Water Concentration'
COC Type
COC
(PPb)
Days of
Exceedance
RQ
Import and repackaging
25,000
39
7.13
825
Acute
16,700
N/A
0.04
724
Chronic
2.8
364
258.57
Incorporation into paints and coatings -
1-part coatings
25,000
57
10.98
1990
Acute
16,700
N/A
0.11
1800
Chronic
2.8
364
642.85
Incorporation into paints and coatings -
2-part reactive coatings
25,000
4
65.90
1240
Acute
16,700
N/A
0.07
1010
Chronic
2.8
364
360.71
Use in paints and coatings - spray
application
25,000
1
2.31
9.1
Acute
16,700
N/A
5.4E-04
7.3
Chronic
2.8
77
2.61
Formulation of TCEP into 2-part
reactive resins
25,000
3
45.51
759
Acute
16,700
N/A
0.04
619
Chronic
2.8
349
221.07
" Production volume of 25,000 lb TCEP/year uses central tendency estimates (50th percentile for all COUs).
b Max day average represents the maximum concentration over a 1- or 30-day average period corresponding with the acute or chronic COC used for the RQ estimate.
"Flow inputs for PSC represent the 50th percentile 7Q10 flows as the lowest expected weekly flow over a 10-year period.
N/A = Days of exceedance are modeled for the application of chronic COCs and do not apply for acute COCs and corresponding RQs.
Page 519 of 638
-------
TableApx H-10. Environmental RQs by Exposure Scenario with Production Volumes of 25,000 lb/year with central tendency release
estimates for Aquatic Organisms with TCEP Surface Water Concentration (ppb) Modeled by VVWM-PSC with 90% percentile flow
of the 7Q10
Exposure
Scenario
Production
Volume (lb/year)"
Days of
Release
Release
(kg/day)
Modeled Using WWM-PSCC
Max Day Ave
(Pl'b)*
COC Type
COC (ppb)
Days of Exceedance
(days per year)
RQ
Import and
repackaging
25,000
39
7.13
9.80
Acute
16,700
N/A
5.90E-04
9.80
Chronic
2.8
51
3.5
Incorporation into
paints and
coatings - 1-part
coatings
25,000
57
10.98
11.40
Acute
16,700
N/A
6.80E-04
11.40
Chronic
2.8
71
4.07
Incorporation into
paints and
coatings - 2-part
reactive coatings
25,000
4
65.90
68.40
Acute
16,700
N/A
4.10E-03
9.12
Chronic
2.8
30
3.25
Use in paints and
coatings at - spray
application
25,000
1
2.31
3.20
Acute
16,700
NA
1.9E-04
0.11
Chronic
2.8
0
0.04
Formulation of
TCEP into 2-part
reactive resins
25,000
3
45.51
7.20
Acute
16,700
N/A
4.30E-04
0.72
Chronic
2.8
0
0.25
" Production volume of 25,000 lb TCEP/year uses central tendency estimates (50th percentile for all COUs).
h Max day average represents the maximum concentration over a 1- or 30-day average period corresponding with the acute or chronic COC used for the RQ estimate.
"Flow inputs for PSC represent the 90th percentile 7Q10 flows as the lowest expected weekly flow over a 10-year period.
N/A = Days of exceedance are modeled for the application of chronic COCs and do not apply for acute COCs and corresponding RQs.
Page 520 of 638
-------
TableApx H-ll. Environmental RQs by Exposure Scenario with Production Volumes of 25,000 lb/year with Central Tendency
Release Estimates for Aquatic Organisms with TCEP Benthic Pore Water Concentration (ppb) Modeled by VVWM-PSC with 90%
Percentile Flow of the 7Q10
Exposure Scenario
Production
Volume
(lb/year)fl
Days of
Release
Release
(kg/day)
Benthic Pore
Water
Concentration
(ppb)6
Benthic Pore Water Concentration'
COC Type
COC
(PPb)
Days of
Exceedance
RQ
Import and repackaging
25,000
39
7.13
4.79
Acute
16,700
N/A
2.90E-04
4.17
Chronic
2.8
51
1.49
Incorporation into paints and coatings -
1-part coatings
25,000
57
10.98
7.11
Acute
16,700
N/A
4.30E-04
6.4
Chronic
2.8
94
2.28
Incorporation into paints and coatings -
2-part reactive coatings
25,000
4
65.90
4.55
Acute
16,700
N/A
2.70E-04
3.6
Chronic
2.8
25
1.29
Use in paints and coatings at - spray
application
25,000
1
2.31
0.05
Acute
16,700
N/A
3.00E-06
0.04
Chronic
2.8
0
1.42E-02
Formulation of TCEP into 2-part
reactive resins
25,000
3
45.51
0.36
Acute
16,700
N/A
2.20E-05
0.29
Chronic
2.8
0
0.10
" Production volume of 25,000 lb TCEP/year uses central tendency estimates (50th percentile for all COUs).
h Max day average represents the maximum concentration over a 1- or 30-day average period corresponding with the acute or chronic COC used for the RQ estimate.
"Flow inputs for PSC represent the 90th percentile 7Q10 flows as the lowest expected weekly flow over a 10-year period.
N/A = Days of exceedance are modeled for the application of chronic COCs and do not apply for acute COCs and corresponding RQs.
Page 521 of 638
-------
H.2 Risk Estimation for Terrestrial Organisms
Table Apx H-12. Calculated RQs Based on TCEP Soils Concentrations (mg/kg) as Calculated
Using Modeled Data for Air Deposition to Soil
Exposure Scenario
Production
Volume (lb/year)fl
Meteorological
Model6
Soil Concentration
(mg/kg) at 1,000 mc
Chronic RQ (Hazard
Value: 612 mg/kg)
Import and Repackaging
2,500
MetCT
1.49E-06
2.43E-09
MetHIGH
1.92E-06
3.14E-09
25,000
MetCT
5.43E-06
8.87E-09
MetHIGH
7.59E-06
1.24E-08
Incorporation into paints
and coatings - 1-part
coatings
2,500
MetCT
3.33E-05
5.44E-08
MetHIGH
5.67E-05
9.27E-08
25,000
MetCT
7.59E-05
1.24E-07
MetHIGH
8.24E-05
1.35E-07
Incorporation into paints
and coatings - 2-part
reactive coatings
2,500
MetCT
1.1 IE—05
1.82E-08
MetHIGH
2.41E-05
3.94E-08
25,000
MetCT
2.19E-05
3.59E-08
MetHIGH
3.68E-05
6.01E-08
Use in paints and coatings
at - spray application
2,500
MetCT
3.97E-03
6.49E-06
MetHIGH
5.58E-03
9.11E-06
25,000
MetCT
5.59E-02
9.14E-05
MetHIGH
8.65E-02
1.41E-04
Formulation of TCEP-
containing reactive resins
(for use in 2-part systems)
2,500
MetCT
3.89E-05
6.35E-08
MetHIGH
3.85E-05
6.30E-08
25,000
MetCT
2.93E-05
4.79E-08
MetHIGH
2.82E-05
4.60E-08
Processing into 2-part
resin article
2,500
MetCT
1.21E-05
1.97E-08
MetHIGH
2.57E-05
4.20E-08
25,000
MetCT
2.71E-05
4.42E-08
MetHIGH
4.58E-05
7.48E-08
Laboratory chemicals
2,500
MetCT
4.84E-05
7.90E-08
MetHIGH
4.65E-05
7.59E-08
25,000
MetCT
2.75E-05
4.50E-08
MetHIGH
2.68E-05
4.37E-08
11 Production volume of 2,500 lb TCEP/yr uses high-end estimates (95th percentile). Production volume of 25,000 lb
TCEP/yr uses central tendency estimates (median).
h The ambient air modeled concentrations and deposition values are presented for two meteorology conditions (Sioux
Falls, South Dakota, for central tendency meteorology; and Lake Charles, Louisiana, for higher-end meteorology).
c Estimated concentrations of TCEP (90th percentile) that could be in soil via air deposition at a community (1,000 m
from the source) exposure scenario.
Page 522 of 638
-------
H.3 Trophic Transfer Analysis Results
Table Apx H-13. RQs Based on Potential Trophic Transfer of TCEP in Terrestrial Ecosystems Using EPA's Wildlife Risk Model for
Eco-SSLs (Equation 4-1)
Exposure Scenario
PV (lb/year)fl
Model6
Soil Concentration
(mg/kg) at 1,000 mc
Nematode
Mammal
Short-Tailed Shrew
TCEP in
biota
(mg/kg/day)
RQ
TCEP in
Biota
(mg/kg/day)
RQ
TCEP in
Biota
(mg/kg/day)
RQ
Import and Repackaging
2,500
MetCT
1.49E-06
1.5E-06
2.4E-09
1.2E-06
2.7E-08
1.2E-06
1.8E-06
MetHIGH
1.92E-06
1.9E-06
3.1E-09
1.5E-06
3.5E-08
1.5E-06
2.3E-06
25,000
MetCT
5.43E-06
5.4E-06
8.9E-09
4.3E-06
9.8E-08
4.3E-06
6.5E-06
MetHIGH
7.59E-06
7.6E-06
1.2E-08
6.0E-06
1.4E-07
6.0E-06
9.1E-06
Incorporation into paints and
coatings - 1-part coatings
2,500
MetCT
3.33E-05
3.3E-05
5.4E-08
2.6E-05
6.0E-07
2.6E-05
4.0E-05
MetHIGH
5.67E-05
5.7E-05
9.3E-08
4.5E-05
1.0E-06
4.5E-05
6.8E-05
25,000
MetCT
7.59E-05
7.6E-05
1.2E-07
6.0E-05
1.4E-06
6.0E-05
9.1E-05
MetHIGH
8.24E-05
8.2E-05
1.3E-07
6.5E-05
1.5E-06
6.5E-05
9.9E-05
Incorporation into paints and
coatings - 2-part reactive
coatings
2,500
MetCT
1.1IE—05
1.1E-05
1.8E-08
8.8E-06
2.0E-07
8.8E-06
1.3E-05
MetHIGH
2.41E-05
2.4E-05
3.9E-08
1.9E-05
4.4E-07
1.9E-05
2.9E-05
25,000
MetCT
2.19E-05
2.2E-05
3.6E-08
1.7E-05
4.0E-07
1.7E-05
2.6E-05
MetHIGH
3.68E-05
3.7E-05
6.0E-08
2.9E-05
6.6E-07
2.9E-05
4.4E-05
Use in paints and coatings-
spray application
2,500
MetCT
0.004
0.004
6.4E-06
0.003
6.8E-05
0.003
0.005
MetHIGH
0.006
0.0056
9.0E-06
0.004
9.8E-05
0.004
0.007
25,000
MetCT
0.056
0.059
9.6E-05
0.044
1.0E-03
0.044
0.067
MetHIGH
0.086
0.086
1.4E-04
0.068
1.5E-03
0.068
0.103
Formulation of TCEP -
containing reactive resins (for
use in 2-part systems)
2,500
MetCT
3.89E-05
3.9E-05
6.4E-08
3.1E-05
7.0E-07
3.1E-05
4.7E-05
MetHIGH
3.85E-05
3.9E-05
6.3E-08
3.1E-05
7.0E-07
3.1E-05
4.6E-05
25,000
MetCT
2.93E-05
2.9E-05
4.8E-08
2.3E-05
5.3E-07
2.3E-05
3.5E-05
MetHIGH
2.82E-05
2.8E-05
4.6E-08
2.2E-05
5.1E-07
2.2E-05
3.4E-05
Page 523 of 638
-------
Exposure Scenario
PV (lb/year)fl
Model6
Soil Concentration
(mg/kg) at 1,000 mc
Nematode
Mammal
Short-Tailed Shrew
TCEP in
biota
(mg/kg/day)
RQ
TCEP in
Biota
(mg/kg/day)
RQ
TCEP in
Biota
(mg/kg/day)
RQ
Processing into 2-part resin
article
2,500
MetCT
1.21E-05
1.2E-05
2.0E-08
9.6E-06
2.2E-07
9.6E-06
1.5E-05
MetHIGH
2.57E-05
2.6E-05
4.2E-08
2.0E-05
4.6E-07
2.0E-05
3.1E-05
25,000
MetCT
2.71E-05
2.7E-05
4.4E-08
2.2E-05
4.9E-07
2.2E-05
3.3E-05
MetHIGH
4.58E-05
4.6E-05
7.5E-08
3.6E-05
8.3E-07
3.6E-05
5.5E-05
Laboratory chemicals
2,500
MetCT
4.84E-05
4.8E-05
7.9E-08
3.8E-05
8.7E-07
3.8E-05
5.8E-05
MetHIGH
4.65E-05
4.6E-05
7.6E-08
3.7E-05
8.4E-07
3.7E-05
5.6E-05
25,000
MetCT
2.75E-05
2.8E-05
4.5E-08
2.2E-05
5.0E-07
2.2E-05
3.3E-05
MetHIGH
2.68E-05
2.7E-05
4.4E-08
2.1E-05
4.8E-07
2.1E-05
3.2E-05
a py = Production volume of 2,500 lb TCEP/yr uses high-end estimates (95th percentile); PV of 25,000 lb TCEP/yr uses central tendency estimates (median).
h The ambient air modeled concentrations and deposition values are presented for two meteorology conditions (Sioux Falls, South Dakota, for central tendency
meteorology; and Lake Charles, Louisiana, for higher-end meteorology).
c Estimated concentrations of TCEP (90th percentile) that could be in soil via air deposition at a community (1,000 m from the source) exposure scenario.
Page 524 of 638
-------
TableApx H-14. RQs Based on Potential Trophic Transfer of TCEP from Fish to American Mink
as a Model Aquatic Pret
ator Using EPA's Wildlife Risk Model for Eco-SS
^s (Equation 4-1)
Scenario Name
Production
Volume
(lb/year)fl
Release
Distribution
SWC6
(^g/L)
Fish
Concentration
(mg/kg)
American Mink
TCEP in Biota
(mg/kg/day)
RQ
Import and repackaging
2,500
High-end
2,370
0.81
0.51
0.02
Incorporation into paints
and coatings - 1-part
coatings
2,500
High-end
10,300
3.50
2.21
0.08
Incorporation into paints
and coatings - 2-part
reactive coatings
2,500
High-end
9,340
3.18
2.01
0.07
Use in paints and coatings
- spray application
2,500
High-end
5,580
1.90
1.20
0.04
Formulation of TCEP
containing reactive resin
2,500
High-end
10,900
3.71
2.34
0.08
Laboratory chemicals
2,500
High-end
96
3.2E-02
0.02
7.0E-04
Import and repackaging
25,000
Central
tendency
1,720
0.58
0.37
0.01
Incorporation into paints
and coatings - 1-part
coatings
25,000
Central
tendency
3,230
1.10
0.69
0.02
Incorporation into paints
and coatings - 2-part
reactive coatings
25,000
Central
tendency
19,300
6.56
4.15
0.14
Use in paints and coatings
- spray application
25,000
Central
tendency
555
0.19
0.12
4.1E-03
Processing into 2-part resin
article
25,000
Central
tendency
15,800
5.37
3.39
0.12
Laboratory chemicals
25,000
Central
tendency
663
0.23
0.14
5.0E-03
11 Production volume of 2,500 lb TCEP/yr uses high-end estimates (95th percentile for all COUs except the laboratory
chemicals COU uses the 1st percentile). Production volume of 25,000 lb TCEP/yr uses central tendency estimates
(median).
b TCEP Surface Water Concentration (SWC) calculated using WWM-PSC.
Page 525 of 638
-------
Appendix I GENERAL POPULATION EXPOSURE DETAILS
1.1 Exposure Factors
Table Apx 1-1. Body Weight by Age Group
Age Group"
Mean Body Weight (kg)A
Infant (<1 year)
7.83
Young toddler (1 to <2 years)
11.4
Toddler (2 to <3 years)
13.8
Small child (3 to <6 years)
18.6
Child (6 to <11 years)
31.8
Teen (11 to <16 years)
56.8
Adults (16 to <78 years)
80.0
11 Age group weighted average
b U.S. EPA (2011a). Table 8-1
Table Apx 1-2. Fish Ingestion
iates by Age Group
Age Group
Fish Ingestion Rate
(g/kg-day)fl
50th Percentile
90th Percentile
Infant (<1 year)6
N/A
N/A
Young toddler (1 to <2 years)b
0.053
0.412
Toddler (2 to <3 years)6
0.043
0.341
Small child (3 to <6 years)6
0.038
0.312
Child (6 to <11 years)6
0.035
0.242
Teen (11 to <16 years)6
0.019
0.146
Adult (16 to <78 years)c
0.063
0.277
Subsistence fisher (adult)'5'
1.78
11 Age group weighted average, using body weight from Table Apx 1-1 above
U.S. EPA (2014a). Table 20a
CU.S. EPA (2014a). Table 9a
1#U.S. EPA (2000b)
Page 526 of 638
-------
1.2 Water Pathway
1.2.1 Surface Water and Groundwater Monitoring Database Retrieval and Processing
The complete set of TCEP monitoring results stored in the WQP was retrieved in March 2023, with no
filters applied other than the chemical name (NWOMC. 2022). This raw dataset included 17,521
samples. To filter down to only the desired surface water samples to include in this analysis, only
samples with the "ActivityMediaSubdivisionName" attribute of "Surface Water" were kept. The dataset
removed values that that were below the detection limit.
After these steps, a total of 466 surface water samples and 51 groundwater samples remained in the
dataset. This monitoring dataset is attached as the Risk Evaluation for Tris(2-chloroethyl) Phosphate
(TCEP) - Supplemental Information File: Water Quality Portal Processed Water Data (U.S. EPA.
2024o).
1.2.1.1 Water Plots and Figures Generated in R
Exploratory analysis of the WQP data were conducted in R. An Rmarkdown file summarizing the steps
taken to explore, wrangle and visualize this dataset is available at (U.S. EPA 2024b).
The Water Media Maps and Time Series Graphs are interactive plots made with the leaflet and plotly
packages. Clicking on the points in the water media maps displays summary information of the
associated data point. Similarly hovering over the data points in the Time Series Graphs provides
summary information of the plotted data point. Media can be selected and de-selected in the legend to
display and remove select media from the figures. The tiles to the left in the media maps allow for
different map layers (Esri.WorldGrayCanvas, OpenStreetMap, Esri.WorldTopoMap) and allows users to
select and deselect the underlying datasets.
Page 527 of 638
-------
Map of Water monitoring in the United States (excludes non-detects)
CANADA
USGS-ID
Value
Units
Location
Type
Fraction
610.00
ug/l
08S 19E 02DCCD1
Well
Dissolved
OrganizationUSGS Idaho Water Science Center
Provider NWIS
Media
Effluent
Finished Water
Groundwater
Hyporheic zone
Landfill effluent
Leachate
Surface Water
Missing
TIT
UNITED STATES
Leaflet | Tiles © Esri — Esri, DeLorme, NAVTEQ
Time Series Graphs
Plot of Water in the United States by Time (excluding non-detects on log scale)
Water
(Apr 2002, 2.78533)
610 ug/l
Idaho
08S 19E 02DCCD1
Well
Dissolved
USGS Idaho Water Science Center
NWIS
Groundwater I ngjFiO I
-------
1.2.2 Methodology for Obtaining New Flow Data (2015 to 2020)
The following steps were utilized to retrieve more recent flow data for the TCEP environmental
assessment (flow values for the 2015 to 2020 are summarized in Risk Evaluation for Tris(2-chloroethyl)
Phosphate (TCEP) - Supplemental Information File: E-FAST Modeling Results (U.S. EPA. 2024g):
1. SIC codes assigned to TCEP were provided: 2851, 4952, 2821, 2823, 2824.
2. Wastewater discharge facility information was obtained for all facilities assigned to each of the
SIC codes using the "echoWaterGetFacilitylnfo" function in the echor package in R. This results
in -47,000 facilities.
3. A data field was added to categorize the SIC codes into new industrial sector names as described
in Table 3 of Versar's "Facility and Stream Flow Database" document. These include "Paint
Formulation," "POTWs—All facilities," and "Adhesives, Sealants, Plastics, Resins, Rubber, and
Manufacturing."
4. For the 4952 SIC code, only facilities with a "POTW" indicator in the permit component data
field were included. This results in a list of-19,000 facilities. This step was taken in parallel to
one described in EPA Contractor Versar's "Facility and Stream Flow Database" document,
where instead of acquiring facilities with a 4,952 SIC designation, all NPDES with a POTW
permit component were retrieved from the water facility search tool in ECHO. Note: Versar also
created a subset "Industrial POTW" category by extracting NPDES permits with a "Y" pre-
treatment indicator from the "POTW—All facilities" category, using the ICI-NPDES database
on the ECHO website.
5. Any duplicate NPDESs were excluded.
6. Four hundred facilities were selected at random without replacement from each industrial sector
group. This step was taken because 19,000 facilities is too many to acquire NHD flow
information for in a timely manner.
7. NHD 14-digit reach codes were retrieved from the ECHO
"dmr_rest_services.get_facility_report" backend server for each unique NPDES/permit that was
active between 2015 to 2020, thus narrowing the facilities to only those with active permits
during this time.
8. Facilities where a NPDES identifier could not be matched with a NHD reach code were
excluded. 877 facilities had active permits during this time period and which also included
reported NHD reach codes.
9. For each unique NPDES-reach code combination, mean and monthly average flow data were
retrieved from the NHD flowline database. Exposure related flow metrics (e.g., 7Q10 and 30Q5)
were then calculated using methods established by the 1,4-D and 1,1-DCA teams.
10. The distribution of flows was plotted.
11. A summary statistics table was created for each of the industrial SIC categories.
1.2.3 E-FAST: Predicted Flowing Surface Water Concentrations (First Tier Modeling)
EPA's E-FAST, Version 2.0, was specifically developed to support EPA assessments of potential
environmental exposures. The E-FAST Model contains default parameter values that allow for exposure
estimations of a chemical in the surface water after a source emits the chemical into a water body
considering simple dilution. EPA uses H-l to estimate surface water concentrations in E-FAST.
Page 529 of 638
-------
EquationApx 1-1.
Where:
swc
R
CF1
T
SF
Surface water concentration in |ig/L
Release kg/site/day
Conversion factor (109 |ig/kg)
Percent removal, typically from wastewater treatment
Flow of receiving river (MLD)
CF2 = Conversion factor (106 L/day/MLD)
Inputs
Release (kg/site/day): As discussed in Section 3.2, the daily release values (kg/site/day) were calculated
using a production volume of 2,500 lb/year, 25,000 lb/yr, emission factors (kg TCEP released/kg TCEP
handled), and number of release days per year. Refer to Table 3-3 for a summary of the release values
by COU, and for sub-scenario-specific release values.
Removalfi'om Wastewater Treatment (%): Removal from wastewater treatment is the percentage of the
chemical removed from wastewater during treatment before discharge to a body of water. Although
removal from wastewater treatment for TCEP was estimated as 0 percent. This is a conservative
estimate relative to what is indicated in Table 2-2 that indicates wastewater removal to be 5 percent for
primary treatment and 19.1 percent for complete treatment (Kim et al.. 2017). EPA assumed that "on-
site WWT," "POTW" release types and direct releases to water did not receive wastewater treatment and
no wastewater treatment removal was applied. This is a conservative assumption that results in the total
amount of TCEP released to wastewater treatment at a direct discharging site being released to surface
water. It reflects the uncertainty of the type of wastewater treatment that may be in use at a direct
discharging facility and the TCEP removal efficiency in that treatment.
Flow of Receiving River (Million L Day): E-FAST requires the selection of a receiving stream flow from
the E-FAST 2014 database. For site-specific assessments, the stream flow is selected by searching for a
facility's NPDES permit number, name, or the known discharging waterbody reach code. As no specific
facilities were identified for the TCEP assessment for water releases, stream flows were selected using
the "SIC Code Option" within E-FAST. This option uses the 10th and 50th percentile stream flows of all
facilities in a given industry sector, as defined by the SIC codes of the industry sector. The associated
SIC Codes for the COU/OES are organized as presented in Table Apx 1-3 below.
Page 530 of 638
-------
Table Apx 1-3. Crosswalk of CPU and PES, Abbreviations, and Relevant SIC Codes
cou
OES
Abbreviation
SIC Code
Manufacturing - Import - Import
Repackaging of import
containers
MFG-IMP
POTW All
Processing - Incorporation into
formulation, mixture, or reaction
product - Flame retardant in: Paint
and coating manufacturing
Incorporation into paints and
coatings - 1-part coatings
PAINT-WB
Paint Formulation
Processing - Incorporation into
formulation, mixture, or reaction
product - Flame retardant in: Paint
and coating manufacturing
Incorporation into paints and
coatings - 2-part reactive
coatings
PAINT-SB
Paint Formulation
Commercial use - Paints and
coatings
Use in paints and coatings at
job sites
COM
POTW All
Processing - Incorporation into
formulation, mixture, or reaction
product - Flame retardant in:
Polymers
Formulation of TCEP
containing reactive resin
PROC
Plastic Resins and
Synthetic Fiber
Manufacture
Use of laboratory chemicals
Wastewater to on-site
treatment or discharge to
POTW (with or without
pretreatment)
LAB
POTW All
These SIC Code stream flows were selected because they were thought to best represent the industrial
activity associated with the COUs and release type.
The flow of rivers is highly variable and is dependent on many factors such as weather patterns and
effluent released from different facilities. The volume of a river varies over time with different flows
expected seasonally and from year to year. The 50th percentile 7Q10 flows represent the lowest
expected weekly flow over a 10-year period and were selected for use in the ecological risk assessment.
The flows for the selected industry sector/SIC Code are shown in Table Apx 1-4. Although not used in
the ecological assessment, harmonic means are also shown because they were used to calculate surface
water concentrations for the scenario specific fish ingestion scenario in the highly exposed human
exposure assessment. Harmonic mean flow values represent long-term average flow conditions.
Page 531 of 638
-------
TableApx 1-4. Harmonic Mean, 30Q5, 7Q10, and 1Q10 50th Percentile Flows for Relevant TCEP
SIC Codes
Sector within E-FAST
Year(s)
Harmonic Mean
Flow MLD
(50th Percentile)
30Q5 Flow
MLD
(50th
Percentile)
7Q10 Flow
MLD
(50th
Percentile)
1Q10 Flow
MLD
(50th
Percentile)
SIC Code - POTW -
All Facilities
2009
1.11E01
1.94
1.06
9.60E-01
2015-2020
1.15E01
7.23
4.13
3.47
SIC Code - Paint
Formulation
2009
3.54E01
1.25E01
7.29
6.10
2015-2020
9.21
5.95
3.38
2.84
SIC Code - Plastic
Resins and Synthetic
Fiber Manufacture
2009
4.45E01
1.37E01
8.02
7.44
2015-2020
6.51
5.05
2.85
2.40
Outputs
The supplemental document entitled Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) -
Supplemental Information File: Exposure EFAST 2014 Surface Water Modeling Inputs, Flow Data, and
General Population Exposure Estimates and Risk Calculations (U.S. EPA. 2024s) provides the inputs,
outputs, and equations that were utilized for calculating surface water concentrations of TCEP, drinking
water estimates, diluted drinking water estimates, incidental oral ingestion estimates from swimming
and incidental dermal absorption estimates from swimming.
Advantages to the E-FAST 2014 model are that it requires minimal input parameters, and it has
undergone extensive peer review by experts outside of EPA. The limitations associated with use of the
E-FAST 2014 model relate to the assumptions made regarding use of sector-based flow information as a
surrogate for site-specific flow information, as well as lack of partitioning (between dissolved and
suspended sediment within the water column or between the water column and the benthic environment)
and degradation parameters that were employed in the PSC model. Additionally, note that low-flow
stream inputs combined with high-release estimates may yield overly conservative surface water
concentrations greater than the water solubility of TCEP.
Page 532 of 638
-------
1.2.3.1 E-FAST 2014 Exposure Activity Parameters
Table Apx 1-5. Incident
tal Derma
(Swimming) Modeling Parameters
Input
Description
(Units)
Adult
(>21
years)
Youth
(11-15
years)
Child
(6-10
years)
Notes
Reference
BW
Body weight (kg)
80
56.8
31.8
EPA Exposure Factors Handbook Chapter
8 (2011), Table 8-1 mean body weight
(U.S. EPA
2011)
SA
Skin surface area
exposed (cm2)
19,500
15,900
10,800
U.S. EPA Swimmer Exposure Assessment
Model (SWIMODEL), 2015
(U.S. EPA
2015)
ET
Exposure time
(hr/day)
3
2
1
High-end default short-term duration from
U.S. EPA Swimmer Exposure Assessment
Model (SWIMODEL), 2015.
(U.S. EPA
2015)
ED
Exposure duration
(years for ADD)
33
5
5
Number of years in age group, up to the
95th percentile residential occupancy
period. EPA Exposure Factors Handbook
Chapter 16(2011), Table 16-5.
(U.S. EPA
2011)
AT
Averaging time
(years for ADD)
33
5
5
Number of years in age group, up to the
95th percentile residential occupancy
period. EPA Exposure Factors Flandbook
Chapter 16(2011), Table 16-5.
(U.S. EPA
2011)
KP
Permeability
coefficient (cm/hr)
2.20E-03
CEM 3.2 estimated aqueous Kp based on
log Kow of 1.25
(Abdallah et
al.. 2016)
Table Apx 1-6. Incidenta
Oral Ingestion i
Swimming) Modeling Parameters
Input
Description
(Units)
Adult
(> 21
years)
Youth
(11-15
years)
Child
(6-10
years)
Notes
Reference
IRmc
Ingestion rate (L/hr)
0.092
0.152
0.096
EPA Exposure Factors Handbook
Chapter 3 (2019), Table 3-7, upper
percentile ingestion while swimming.
(U.S. EPA
2019)
BW
Body weight (kg)
80
56.8
31.8
EPA Exposure Factors Handbook
Chapter 8 (2011), Table 8-1 mean body
weight.
(U.S. EPA
2011)
ET
Exposure time
(hr/day)
3
2
1
High-end default short-term duration
from U.S. EPA Swimmer Exposure
Assessment Model (SWIMODEL),
2015; based on competitive swimmers
in the age class.
(U.S. EPA
2015)
IRmc-
daily
Incidental daily
ingestion rate
(L/day)
0.276
0.304
0.096
Calculation: ingestion rate x exposure
time
IR/BW
Weighted incidental
daily ingestion rate
(L/kg-day)
0.0035
0.0054
0.0030
Calculation: ingestion rate/body weight
ED
Exposure duration
(years for ADD)
33
5
5
Number of years in age group, up to the
95th percentile residential occupancy
period. EPA Exposure Factors
(U.S. EPA
2011)
Page 533 of 638
-------
Input
Description
(Units)
Adult
(> 21
years)
Youth
(11-15
years)
Child
(6-10
years)
Notes
Reference
Handbook Chapter 16 (2011), Table 16-
5.
AT
Averaging time
(years for ADD)
33
5
5
Number of years in age group, up to the
95th percentile residential occupancy
period. EPA Exposure Factors
Handbook Chapter 16 (2011), Table 16-
5.
(U.S. EPA.
2011)
CF1
Conversion factor
(mg/jig)
1.00E-03
CF2
Conversion factor
(days/year)
365
1.2.4 VVWM-PSC: Predicted Flowing Surface Water Concentrations (Second Tier
Modeling)
Site-specific parameters influence how partitioning occurs over time. For example, the concentration of
suspended sediments, water depth, and weather patterns all influence how a chemical may partition
between compartments. Physical and chemical properties of the chemical itself also influence
partitioning and half-lives into environmental media. TCEP has a Koc greater than 100, indicating a high
potential to sorb to suspended particles in the water column and settled sediment in the benthic
environment.
EPA conducted higher tier modeling with PSC-VVWM to estimate benthic concentrations (porewater
and sediment).
1.3 Ambient Air Pathway
This section provides an overview of EPA's screening-level methodology for the ambient air pathway.
Where reasonably available, fugitive and stack air release data from the 2019 TRI are used to quantify
environmental releases. No TRI data were available for TCEP. EPA used estimated releases from a
hypothetical facility using TCEP for the COUs (Figure Apx 1-2).
AERMOD is used to estimate ambient air concentrations and exposures to human populations at various
distances from the emission source. Distances of up to 10,000 m are evaluated to capture potential
exposures and associated risks to fenceline communities. A distance of 10,000 m is used for this
methodology to capture populations nearer to releasing facilities than may otherwise be evaluated under
other EPA administered laws. Additionally, professional knowledge and experience regarding exposures
associated with the ambient air pathway find risks frequently occur out to approximately 1,000 m from a
releasing facility and quickly decrease farther out. Although 10,000 m is an order of magnitude farther
out than where risks are expected to occur, 10,000 m provides an opportunity to capture other factors
related to potential exposure and associated potential risks via the ambient air pathway (like multiple
facilities impacting a single individual) providing flexibility for screening-level analyses for future risk
evaluations. While 10,000 m is used for the outer distance in the screening-level analysis, the
methodology is not limited to 10,000 m. If risks are identified out to 10,000 m, then additional analysis
using the screening-level methodology can be extended to farther distances for purposes of identifying
where risks may fall below levels of concern.
Page 534 of 638
-------
• Stack and • Ambient air ~ Inhalation • Fenceline
fugitive air concentrations Community
releases from
iv ^
Source Pathway Route ^ Receptors
modeling
Distance
from
Source
FigureApx 1-2. Overview of EPA's Screening-Level Ambient Air Pathway Methodology
1.3.1 Modeling Approach for Estimating Concentrations in Ambient Air
EPA applied a tiered approach to estimate ambient air concentrations and exposures for members of the
general population that are in proximity (between 10 to 10,000 m) to emissions sources emitting the
chemicals being evaluated to the ambient air. All exposures were assessed for the inhalation route only.
For TCEP, multi-year release data were not available.
Step 1: Ambient Air: IIOAC Methodology
Methodology is scenario-specific. Analysis evaluates ambient air concentrations and associated
exposures/risks resulting from facility-specific releases at three pre-defined distances (100, 100
to 1,000, and 1,000 m) from a releasing facility.
Step 2: Ambient Air: AERMOD Methodology
Methodology is scenario-specific. Analysis evaluates ambient air concentrations and associated
exposures/risks, and deposition concentrations to land and water, resulting from facility-specific
releases at eight finite distances (10, 30, 60, 100, 1,000, 2,500, 5,000, and 10,000 m) and two
area distances (30 to 60 m and 100 to 1,000 m) from each releasing facility (or generic facility
for alternative release estimates).
1.3.2 Ambient Air: Screening Methodology
The Ambient Air: IIOAC Methodology identifies, at a high level, if there are inhalation exposures to
select human populations from a chemical undergoing risk evaluation that indicates a potential risk. This
methodology inherently includes both estimates of exposures as well as estimates of risks to inform the
need, or potential need, for further analysis. If findings from the Ambient Air: IIOAC Methodology
indicate any potential risk (acute non-cancer, chronic non-cancer, or cancer) for a given chemical above
(or below as applicable) typical Agency benchmarks, EPA generally will conduct a higher tier analysis
of exposures and associated risks for that chemical. If findings from the Ambient Air: IIOAC
Methodology do not indicate any potential risks for a given chemical above (or below as applicable)
typical agency benchmarks, EPA would not expect a risk would be identified with higher tier analyses,
but may still conduct a limited higher tier analysis at select distances to ensure potential risks are not
missed (e.g., at distances <100 m to ensure risks do not appear very near a facility where populations
may be exposed).
Page 535 of 638
-------
Model
EPA's 110AC model45 was used to estimate high-end and central tendency (mean) exposures to select
human populations at three pre-defined distances from a facility releasing a chemical to the ambient air
(100, 100 to 1,000, and 1,000 m). IIOAC is a spreadsheet-based tool that estimates indoor and outdoor
air concentrations using pre-run results from a suite of dispersion scenarios run in a variety of
meteorological and land-use settings within EPA's AERMOD. As such, IIOAC is limited by the
parameterizations utilized for the pre-run scenarios within AERMOD (meteorologic data, stack heights,
distances, populations, etc.) and any additional or new parameterization would require revisions to the
model itself. Readers can learn more about the IIOAC model, equations within the model, detailed input
and output parameters, pre-defined scenarios, default values used, and supporting documentation by
reviewing the IIOAC users guide (U.S. EPA 2019f).
Releases
EPA modeled exposures for the following list of COUs/OES that had air releases. EPA ran two
scenarios for each release scenario:
1. Central Tendency (50th percentile) Estimate for High Production Volume (25,000 lb) - HIGH-
CT; and
2. High-End (95th percentile) Estimate for Low Production Volume (2,500 lb) - LOW-HE.
TableApx 1-7. Ambient Air Release Inputs Utilized for Ambient Air Modeling: IIOAC and
AERMOD Methodology for TCEP
Scenario Name
Production
Volume
Estimate
Fugitive/
Stack
Release Duration
(hours/day)
Release
Frequency
(days/year)
Release
Amount
(kg/site/day)
COM-Paints-USE
LOW
HE
Fugitive
8 h/day (8-4 p.m.)
2
1.14E02
IND-LabChem-USE
LOW
HE
Fugitive
8 h/day (8-4 p.m.)
235
2.32E-04
IND-LabChem-USE
LOW
HE
Stack
8 h/day (8-4 p.m.)
235
2.32E-04
MFG-Repack
LOW
HE
Fugitive
1 h/day (12-1 p.m.)
4
3.43E-04
MFG-Repack
LOW
HE
Stack
1 h/day (1 p.m.)
4
3.43E-04
PROC-Article-PROC-
twopart-resin
LOW
HE
Fugitive
8 h/day (8-4 p.m.)
109
4.22E-04
PROC-Article-PROC-
twopart-resin
LOW
HE
Stack
8 h/day (8-4 p.m.)
109
4.22E-04
PROC-Paints-INC-2-part
reactive coatings
LOW
HE
Fugitive
8 h/day (8-4 p.m.)
1
7.90E-03
PROC-Paints-INC-2-part
reactive coatings
LOW
HE
Stack
8 h/day (8-4 p.m.)
1
1.99E-02
PROC-Paints-INC-1 -part
LOW
HE
Fugitive
8 h/day (8-4 p.m.)
4
9.60E-03
PROC-Paints-INC-1 -part
LOW
HE
Stack
8 h/day (8-4 p.m.)
4
9.60E-03
PROC-Polymer-F ORM-
reactive-resin
LOW
HE
Fugitive
8 h/day (8-4 p.m.)
1
8.83E-03
45 The IIOAC website is available at https://www.epa.gov/tsca-screening-tools/iioac-integrated-indoor-outdoor-air-calculator.
Page 536 of 638
-------
Scenario Name
Production
Volume
Estimate
Fugitive/
Stack
Release Duration
(hours/day)
Release
Frequency
(days/year)
Release
Amount
(kg/site/day)
PROC-Polymer-F ORM-
reactive-resin
LOW
HE
Stack
8 h/day (8-4 p.m.)
1
2.07E-02
COM-Paints-USE
HIGH
CT
Fugitive
8 h/day (8-4 p.m.)
1
1.23E01
IND-LabChem-USE
HIGH
CT
Fugitive
8 h/day (8-4 p.m.)
230
1.35E-04
IND-LabChem-USE
HIGH
CT
Stack
1 h/day (1 p.m.)
230
1.35E-04
MFG-Repack
HIGH
CT
Fugitive
1 h/day (12-1 p.m.)
39
1.88E-04
MFG-Repack
HIGH
CT
Stack
1 hr/day (1 p.m.)
39
1.88E-04
PROC-Article-PROC-
twopart-resin
HIGH
CT
Fugitive
8 h/day (8-4 p.m.)
231
1.43E-04
PROC-Article-PROC-
twopart-resin
HIGH
CT
Stack
8 h/day (8-4 p.m.)
231
1.43E-04
PROC-Paints-INC-2-part
reactive coatings
HIGH
CT
Fugitive
8 h/day (8-4 p.m.)
4
6.77E-03
PROC-Paints-INC-2-part
reactive
HIGH
CT
Stack
8 h/day (8-4 p.m.)
4
5.63E-03
PROC-Paints-INC-1 -part
HIGH
CT
Fugitive
8 h/day (8-4 p.m.)
52
1.63E-03
PROC-Paints-INC-1 -part
HIGH
CT
Stack
8 h/day (8-4 p.m.)
52
1.63E-03
PROC-Polymer-F ORM-
reactive-resin
HIGH
CT
Fugitive
8 h/day (8-4 p.m.)
6
5.36E-03
PROC-Polymer-F ORM-
reactive-resin
HIGH
CT
Stack
8 h/day (8-4 p.m.)
8
3.72E-03
Exposure Scenarios
EPA modeled exposure scenarios for two source types: stack (point source) and fugitive (area source)
releases. These source types have different plume and dispersion characteristics accounted for
differently within the IIOAC model. All COUs had stack and fugitive emissions except for the
commercial use of paints and coatings (COM-Paints-USE).
The topography represents an urban or rural population density and certain boundary layer effects (like
heat islands in an urban setting) that can affect turbulence and resulting concentration estimates at
certain times of the day. EPA ran both urban and rural population density for all scenarios.
IIOAC includes 14 pre-defined climate regions (each with a surface station and upper-air station).
Because release data used for the Ambient Air: IIOAC Methodology was not facility- or location-
specific, EPA selected 1 of the 14 climate regions to represent a high-end (South [Coastal]) climate
region. This selection was based on a sensitivity analysis of the average concentration and deposition
predictions. This climate regions selected represents the meteorological dataset that tended to provide
high-end concentration estimates relative to the other stations within IIOAC. The meteorological data
within the IIOAC Model are from years 2011 to 2015 as that is the meteorological data utilized in the
suite of pre-run AERMOD exposure scenarios during development of the IIOAC model (see (U.S. EPA
Page 537 of 638
-------
2019f)). While this is older meteorological data, sensitivity analyses related to different years of
meteorological data found that although the data does vary, the variation is minimal across years so the
impacts to the model outcomes remain relatively unaffected.
The release scenarios were informed by the release duration and release frequency that were provided in
Section 3.2.
Results
The supplemental document entitled Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) -
Supplemental Information File: IIOAC Modeling Inputs and Results (U.S. EPA. 20241) presents the
overall inputs and outputs for IIOAC. In IIOAC, all calculated air concentrations of fine and coarse
particles are capped by an upper limit equal to the National Ambient Air Quality Standards (NAAQS)
for particulate matter (PM) (U.S. EPA. 2016c). These limits are 35 and 150 [j,g/m3 for fine and coarse
particles (i.e., the NAAQS for PM2.5 and PM10), respectively. For the IIOAC results, these limits were
met for all the COU/OES releases with stack emissions. In addition, this limit reach was reached for the
fine, fugitive emissions, LOW-HE release scenario for the commercial use of paints and coatings.
A further limitation of IIOAC is that it does not model for gaseous deposition. Due to the inability to
model gaseous deposition, and due to the initial screening results meeting the NAAQS caps, EPA
decided to run a higher tier model (AERMOD) for the ambient air pathway.
1.3.3 Ambient Air: AERMOD Methodology
The Ambient Air: AERMOD Methodology was developed to allow EPA to conduct a higher tier
analysis of releases, exposures, and associated risks to human populations around releasing facilities at
multiple distances when EPA has site-specific data like reported releases, facility locations (for local
meteorological data), source attribution, and other data when reasonably available. This methodology
can also incorporate additional site-specific information like stack parameters (stack height, stack
temperature, plume velocity, etc.), building characteristics, release patterns, different terrains, and other
parameters when reasonably available. AERMOD can be performed independent of the Tier 1 modeling
described above, provides a more thorough analysis, can include wet and dry deposition estimates, and
allows EPA to fully characterize identified risks for chemicals undergoing risk evaluation. The
application of this methodology can be applied to single or multiple years of data. TCEP had no TRI or
NEI data. Thus, air releases from the release assessment were used to estimated ambient air
concentrations for a single year.
Model
The Ambient Air: AERMOD Methodology for this risk evaluation utilizes AERMOD to estimate TCEP
exposures to fenceline communities at user defined distances from a facility releasing TCEP. AERMOD
is a steady-state Gaussian plume dispersion model that incorporates air dispersion based on planetary
boundary layer turbulence structure and scaling concepts, including treatment of both surface and
elevated sources and both simple and complex terrain. AERMOD can incorporate a variety of emission
source characteristics, chemical deposition properties, complex terrain, and site-specific hourly
meteorology to estimate air concentrations and deposition amounts at user-specified population
distances and at a variety of averaging times. Readers can learn more about AERMOD, equations within
the model, detailed input and output parameters, and supporting documentation by reviewing the
AERMOD Users Guide (U.S. EPA. 2018).
Page 538 of 638
-------
Releases
EPA modeled exposures using the release data developed as described in Section 3.2. Release data were
provided (and modeled) on a COU-by-COU basis as no facility information was available for TCEP.
Exposure Points
The Ambient Air: AERMOD Methodology evaluated exposures to exposure points at eight finite
distances (10, 30, 60, 100, 1,000, 2,500, 5,000, and 10,000 m) and two area distances (30 to 60 m, and
100 to 1,000 m) from each releasing facility (or generic facility for alternative release estimates).
Exposure points for each of the eight finite distances were placed in a polar grid every 22.5 degrees
around the respective distance ring. This results in a total of 16 exposure points around each finite
distance ring for which exposures are modeled. FigureApx 1-3 provides a visual depiction of the
placement of exposure points around a finite distance ring. Although the visual depiction only shows
exposure points locations around a single finite distance ring, the same placement of exposure points
occurred for all eight finite distance rings.
Receptor Locations around each Finite Distance Ring
100 -1,000 m
2,500 m
10,000 m
22.5 '
Releasing Facility
Figure Apx 1-3. Modeled Exposure Points Locations for Finite Distance Rings
Exposure points for the area distance 30 to 60 m evaluated were placed in a cartesian grid at equal
distances between 40 and 50 m around each releasing facility (or generic facility for alternative release
estimates) were placed at 10-meter increments.
Exposure points for the area distance 100 to 1,000 m evaluated were placed in a cartesian grid at equal
distances between 200 and 900 m around each releasing facility (or generic facility for alternative
release estimates) were placed at 100-meter increments. This results in a total of 456 exposure points for
which exposures are modeled. Figure Apx 1-4 provides a visual depiction of the placement of exposure
points (each dot) around the 100 to 1,000 m area distance ring.
Page 539 of 638
-------
FigureApx 1-4. Modeled Exposure Points for Area Distance
All exposure points were at 1.8 m above ground, as a proximation for breathing height for ambient air
concentration estimations. A duplicate set of exposure points was at ground level (0 m) for deposition
estimations.
Meteorological Data
Meteorological data for EPA estimated releases (where TRI or city data were not available) were
modeled with the two meteorological stations utilized in the pre-screen methodology (Sioux Falls, South
Dakota, for central tendency meteorology; Lake Charles, Louisiana, for higher-end meteorology). These
two meteorological stations represent meteorological datasets that tended to provide high-end and
central tendency concentration estimates relative to the other stations within IIOAC based on a
sensitivity analysis of the average concentration and deposition predictions conducted in support of
IIOAC development. These two meteorological stations are based on 5 years of meteorological data
(2011 to 2015) and provide high-end and central tendency exposure concentrations utilized for risk
calculation purposes to identify potential risks. The "ADJ U*" option was not used for the 2011 to 2015
data as this could lead to model overpredictions of ambient concentrations during those particular
conditions.
All processing also used automatic substitutions for small gaps in data for cloud cover and temperature.
Urban/Rural Designations
Urban/rural designations of the area around a facility are relevant when considering possible boundary
layer effects on concentrations.
Air emissions taking place in an urbanized area are subject to the effects of urban heat islands,
particularly at night. When sources are set as urban in AERMOD, the model will modify the boundary
layer to enhance nighttime turbulence, often leading to higher nighttime air concentrations. AERMOD
uses urban-area population as a proxy for the intensity of this effect.
Where TRI or city data were not available for a facility requiring modeling, there was no way for EPA
to determine an appropriate urban or rural designation. Instead, EPA modeled each such facility once as
Page 540 of 638
-------
urban and once as not urban.46 There is no recommended default urban population for AERMOD
modeling, so for these facilities EPA assumed an urban population of 1 million people, which is
consistent with the estimated populations used with IIOAC. Although slightly higher, the assumed urban
population is close to the average of all the urban populations used for the TRI reporting facilities
(which was 847,906 people).
For the TCEP risk evaluation EPA selected the urban air concentrations vs. rural air concentrations as
urban concentrations were generally more conservative. Rural air concentrations may be relevant for
facilities located in rural areas, and because TCEP has long range transport potential. However due to
lack of site-specific information for facilities, this risk evaluation used the more conservative urban air
estimates from AERMOD.
Physical Source Specifications for Alternative Release Estimates
EPA estimated releases (where TRI or city data were not available) were modeled centering all
emissions on one location and using IIOAC default physical parameters. Stack emissions were modeled
from a point source at 10 meters above ground from a 2-meter inside diameter, with an exit gas
temperature of 300 Kelvin and an exit gas velocity of 5 m/sec (see Table 6 of the IIOAC User Guide).
Fugitive emissions were modeled at 3.05 m above ground from a square area source of 10 m on a side
(see Table 7 of the IIOAC User Guide).
Deposition Parameters
AERMOD was used to model daily (g/m2/day) and annual (g/m2/year) deposition rates from air to land
and water at eight finite distances (10, 30, 60, 100, 1,000, 2,500, 5,000, and 10,000 m) and two area
distances (30 to 60 m and 100 to 1,000 m) from each releasing facility (or generic facility for alternative
release estimates).
AERMOD can model both gaseous and particle deposition. For TCEP, EPA considered both gaseous
and particle deposition. There is conflicting literature on whether TCEP is present in particulates vs. gas.
Section 3.3.1.2.1 discusses these differences. Input parameter values for AERMOD deposition modeling
are shown in TableApx 1-8.
EPA provided the parameter values and settings for AERMOD deposition modeling, as indicated in
Table Apx 1-8 and Table Apx 1-9. The particle deposition utilized the "METHOD2" option in
AERMOD, which is recommended when particle size distributions are not well known and when less
than 10 percent of particles (by mass) are 10 |im or larger. Note that we modeled each scenario twice—
once with gaseous deposition utilizing land cover of "suburban area, forested" and once with "bodies of
water."
46 Although this may be viewed as a potential double counting of these releases, EPA only utilized the highest estimated
releases from a single exposure scenario from the suite of exposure scenarios modeled for surrogate/estimated facility
releases as exposure estimates and for associated risk calculations.
Page 541 of 638
-------
Table Apx 1-8. Settings for Gaseous Deposition
Parameter
Value
Source
Diffusivity in air
5.67E-02 cm2/sec
Utilizing www.envmodels.com with the chemical properties
from Table 1 of Shin et al. (2014)
Diffusivity in
water
2.70E-05 cm2/sec
Paae 2310 of Melnikova et al. (2019)
Henry's Law
constant
2.95E-06 Pa m3/mol
Not specified
rci: Cuticular
resistance to
uptake by lipids
for individual
leaves
3.26E03 sec/cm
Based on vapor pressure (Vp = 8.13 Pa), empirical
relationships described bv Welke et al. (1998) and Kerler and
Schoenherr (1988) and the values of rci and of Vp available
for numerous chemicals in Weselv et al. (2002) —toaether.
these imply a relationship of log(rcl) = 0.4892*log(Vp in Pa)
+ 3.0682 "
Seasons
DJF = winter with no snow;
MAM = transitional spring
with partial green coverage
or short annuals; JJA =
Midsummer with lush
vegetation; SON. = Autumn
with unharvested cropland
Assumption
Land Cover
Option 1: Suburban areas,
forested; Option 2: Bodies
of water
A limited set of AERMOD tests suggested suburban-forest
was a reasonable and appropriately health-protective default
land-cover selection when land-cover analysis is not
possible. Bodies of water typically led to the highest
deposition values (ICF unpublished data).
Pa = Pascal; mol = mole; DJF = December-February; MAM = March-May; JJA = June-August; SON =
September-November.
Table Apx 1-9. Setl
tings for Particle Deposition
Parameter
Value
Source
Mass fraction 2.5
|im or smaller
0.4 |im
Based on ranees found for phosphates in (Delumvea
and PeteL 1979) and (Lee and Patterson, 1969)
Mass-mean
diameter
2.2 |im
Based on a default for phosphates (source not
specified)
Cuticular Resistance
The cuticular resistance (rci) value represents the resistance of a chemical to uptake by individual leaves
in a vegetative canopy. For TCEP, rci was not readily available in literature. For chemicals for which the
rci value is not readily available in literature, EPA developed three methods to estimate the rci value. For
TCEP, EPA used rci value estimated using Method 2.
Method 1: Approximation of Rci Value as a Function of Vapor Pressure: Data from the literature
indicate that rci value varies as a function of the vapor pressure (VP, units of Pa) of a chemical (Welke et
al.. 1998; Kerler and Schoenherr. 1988). A high VP indicates that chemical has a high propensity for the
vapor phase relative to the condensed phase, and therefore, would have high resistance to uptake from
Page 542 of 638
-------
the atmosphere into leaves (i.e., high rci). Furthermore, Wesely et al. (2002) provides a large database of
VP and rci values.
Analysis of the that study data (Welke et al.. 1998) reveals that there is a linear correlation between
log(VP) and log(rci), as illustrated in FigureApx 1-5 and EquationApx 1-2. The unit of VP is in Pascals
(Pa). Linear regression yields rci as a function of VP (R2 = 0.606):
EquationApx 1-2.
log(rcl) = 0.489 log (VP) + 3.068
rcl = 1170 VP0A98
16
14
12
10
8
o
4
2
0
-2
-4
-12 -10 -8 -6 -4 -2 0 2 4 6 8
log (VP (Pa))
Figure Apx 1-5. Cuticular Resistance as a Function of Vapor Pressure
Method 2: Empirical Calculation of Cuticular Resistance: Method 2 estimates rci value using various
empirical equations found in literature. This method assumes the vapor pressure of the chemical at 20 to
25 °C is equal to the saturation vapor pressure. For VOCs, using the equations collectively provided
under Equation Apx 1-3 (Welke et al.,) the polymer matrix-air partition coefficient (K\[Xa) can be
calculated as follows:
EquationApx 1-3.
log(Kmxo) = 6.290 - 0.892 log(7P)
Next, KMxa can be converted to the cuticular membrane-air partition coefficient, Kcma:
Kcmci = 0.77 KMXa
Welke, et al. also provide an empirical relationship between the polymer matric-water partition
coefficient and the air-water partition coefficient, Kmxw. Recognizing the air-water partition coefficient
is the Henry's law constant, HLC (unitless), yields,
Kmxw = KMX a HLC
y = 0.4892x +3.0682
R2 = 0.6058
. * ^
•• fir *
• • * ••JV "v"
% ,
Page 543 of 638
-------
This relationship can be generalized from the polymer matrix to the cuticular membrane.
KcMw — KcMa HLC
In a separate study, Kerler and Schoenherr (1988) have developed an empirical relationship that equates
Kcmw to the permeance coefficient for cuticular membranes, Pcm. However, this relationship was
developed using data for non-volatile chemicals. Consequently, applying it to volatile organic chemicals
introduces a large amount of uncertainty to the analysis and may not be scientifically justifiable.
In the above equation, MV is the molecular volume of the chemical in question, which can be calculated
from the molar mass, m (units of g/mol), and density, d (units of g/cm3):
Finally, rci is understood to be the inverse of Pcm. The above relationships can be put together and
simplified to yield a single equation for rcl as a function of vapor pressure, molar mass, and density:
Method 3: Read-Across of Cuticular Resistance from an Analog: This method assumes that chemicals
that have structural similarity, physical and chemical similarity, and exhibit similar vapor pressures will
also exhibit similar rci values. Available data in literature (Wesely et al.. 2002) can be used as a
crosswalk for read-across determination of rci. The unknown rci value is then assumed to be equal to the
rci of the analog.
Ambient Air Exposure Concentration Outputs
Hourly-average concentration outputs were provided from AERMOD for each exposure points around
each distance ring (each of 16 exposure points around a finite distance ring or each exposure points
within the area distance ring). Daily and period averages were then calculated from the modeled hourly
data. Daily averages for the finite distance rings were calculated as arithmetic averages of all hourly data
for each day modeled for each v around each ring. Daily averages for the area distance ring were
calculated as the arithmetic average of the hourly data for each day modeled across all exposure points
within the area distance ring. This results in the following number of daily average concentrations at
each distance modeled.
1. Daily averages for EPA estimated releases: Average concentrations for each of 365 (or 366) days
for each of 16 exposure points around each finite distance ring.
Period averages were calculated from all the daily averages for each exposure points for each distance
ring over 1 year for facilities where releases were estimated. This results in a total of 16 period average
concentration values for each finite distance ring. This is derived from either averaging the daily
averages across the single year of meteorological data used for TRI reporting facilities or across the
multi-year meteorological data used for EPA estimated releases.
-238 d
Page 544 of 638
-------
Daily and period average Outputs were stratified by different source scenarios, such as urban/not urban
setting or emission-strengths where needed. Outputs from AERMOD are provided in units of
micrograms per cubic meter (|ig/m3) requiring conversion to parts per million (ppm) for purposes of
calculating risk estimates for 1,4-dioxane. The following formula was used for this conversion:
EquationApx 1-4.
CPPm= (24.45*(Caermod)/1,000)/MW
Where:
Cppm
24.45
Caermod
MW
Concentration (ppm)
Molar volume of a gas at 25 °C and 1 atmosphere pressure
Concentration from AERMOD (|ig/m3)
Molecular weight of the chemical of interest (g/mole)
Post-processing scripts were used to extract and summarize the output concentrations for each facility,
release, and exposure scenario. The following statistics for daily- and period-average concentrations
were extracted or calculated from the results for each of the modeled distances (i.e., each ring or grid of
exposure points) and scenarios:
• Minimum;
• Maximum;
• Average;
• Standard deviation; and
• 10th, 25th, 50th, 75th, and 95th percentiles.
Table Apx 1-10. Description of Daily or Period Average and Air Concentration Statistics
Statistic
Description
Minimum
The minimum daily or period average concentration estimated at any exposure point on
any day at the modeled distance.
Maximum
The maximum daily or period average concentration estimated at any exposure point on
any day at the modeled distance.
Average
Arithmetic mean of all daily or period average concentrations estimated at all exposure
points locations on all days at the modeled distance. This incorporates lower values
(from days when the exposure point largely was upwind from the facility) and higher
values (from days when the exposure point largely was downwind from the facility).
Percentiles
The daily or period average concentration estimate representing the numerical percentile
value across the entire distribution of all concentrations at all exposure point locations
on any day at the modeled distance. The 50th percentile represents the median of the
daily or period average concentration across all concentration values for all exposure
point locations on any day at the modeled distance.
Deposition from Ambient Air to Soil and Water Exposure Concentration Outputs
As previously mentioned, AERMOD was used to model daily (g/m2/day) and annual (g/m2/year)
deposition rates (i.e., deposition flux) from air releases to water body catchment areas. EPA
quantitatively evaluated the risk to aquatic (pelagic and benthic) and terrestrial organisms from exposure
to soil, surface water bodies and sediment via air deposition resulting from the manufacturing,
processing, use, or disposal of TCEP. The following equations and parameters are based on the generic
Page 545 of 638
-------
farm pond scenario from models, such as the GENEEC2 (Generic Estimated Environmental
Concentration) and EXAM (Exposure Analysis Modeling System) used by EPA. Total deposition for
each media (soil, water body, and sediment) were derived using the deposition rate modeled by
AERMOD to calculate media (soil, water body, and sediment) concentrations using the generic farm
pond parameters for area, mixing depths, and densities, respectively:
Soil:
EquationApx 1-5.
T otal Deposition to Soil Catchment (ug) = Deposition flux x Area x CF
Where:
Deposition flux = Annual deposition flux to water body catchment (g/m2)
Area = Area of soil catchment (area of water body catchment - area of
water body) or 100,000 m2 - 10,000 m2 = 90,000 m2
CF = g to [j,g; 1,000,000
(ug\ (Total Deposition to Soil Catchent)
Soil Catchment Concentration -— = — — ; :—; ; — :—-
\kgj (Area of soil catchment x mix depth x soil density)
Where:
Area = 90,000 m2
Mix depth = 0.1m
Soil density = 1,700 kg/m3
Water Body:
EquationApx 1-6.
Total Deposition to Water Body (ug) = Deposition flux x Area x CF
Where:
Deposition flux = Annual deposition flux to water body catchment (g/m2)
Area = Area of water body; 10,000 m2
CF = g to ug; 1,000,000
ug\ Total Deposition to Water Body
/ug\
Water Body Concentration (— J =
L ) (Area x Pond Depth x CF)
Where:
Area = area of water body; 10,000 m2
Pond depth = 2 m
CF = m3 to L; 1,000
Sediment:
EquationApx 1-7.
Total Deposition to Water Body
Sediment Concentration
(ug\ =
\ka)
\kg) (Area x mix depth x sediment density)
Where:
Area = Area of water body; 10,000 m2
Page 546 of 638
-------
0.1 m
1,300 kg/m3
AERMOD Air Concentrations and Deposition Results
Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File: Exposure
Air Concentration Risk Calculations (U.S. EPA. 2024i) includes the ambient air concentrations,
deposition concentrations (soil, water body, and sediment) for all OESs, and the associated risk
calculations.
1.4 Fish Ingestion Pathway
1.4.1 Exposure Estimates
TableApx 1-11. Adult General Population Fish Ingestion Doses by Scenario Based on a
Production Volume of 2,500 lb/year, High-End Release Distribution, and Modeled Surface Water
Concentrations Based on 90th Percentile Flow of Harmonic Mean
Scenario
Name
swca
(Hg/L)
ADR4
(mg/kg-day)
ADD4 (mg/kg-day)
LADD4 (mg/kg-day)
BAF
2,198
BAF
109
BAF
2,198
BAF
109
BAF
2,198
BAF
109
HE
CT
HE
CT
HE
CT
HE
CT
HE
Import and
Repackaging
5.5
3.38E-03
1.68E-04
7.68E-04
3.38E-03
3.81E-05
1.68E-04
6.11E-04
2.69E-03
3.03E-05
1.33E-04
Incorporation
into Paints
and Coatings
- 1-Part
Coatings
15.4
9.39E-03
4.66E-04
2.14E-03
9.39E-03
1.06E-04
4.66E-04
1.70E-03
7.46E-03
8.42E-05
3.70E-04
Incorporation
into Paints
and Coatings
- 2-Part
Reactive
Coatings
14.0
8.51E-03
4.22E-04
1.94E-03
8.51E-03
9.60E-05
4.22E-04
1.54E-03
6.77E-03
7.63E-05
3.35E-04
Use in Paints
and Coatings
at Job Sites
13.1
7.95E-03
3.94E-04
1.81E-03
7.95E-03
8.97E-05
3.94E-04
1.44E-03
6.32E-03
7.13E-05
3.13E-04
Formulation
ofTCEP
Containing
Reactive
Resin
2.1
1.29E-03
6.37E-05
2.92E-04
1.29E-03
1.45E-05
6.37E-05
2.32E-04
1.02E-03
1.15E-05
5.07E-05
Laboratory
Chemicals
77.1
4.69E-02
2.33E-03
1.07E-02
4.69E-02
5.29E-04
2.33E-03
8.48E-03
3.73E-02
4.21E-04
1.85E-03
" Surface water concentrations based on 90th percentile flow of harmonic mean flow conditions.
4 ADR calculated using the 90th percentile fish ingestion rate (22.2 g/day). ADD and LADD were calculated using both the mean
(CT) and 90th percentile (HE) fish ingestion rates, 5.04 g/day and 22.2 g/day respectively. An ADD based on the 90th percentile
ingestion rate is the same as an ADR.
Page 547 of 638
-------
TableApx 1-12. Adult Subsistence Fisher Doses by Scenario Based on a Production Volume of
2,500 lb/year, High-End Release Distribution, and Modeled Surface Water Concentrations Based
on 90th Percentile Flow of Harmonic Mean
Scenario Name
SWC"
(Mg/L)
ADD, ADR
(mg/kg-day)
BAF 2,198
ADD, ADR
(mg/kg-
day)
BAF 109
LADD
(mg/kg-
day)
BAF 2,198
LADD
(mg/kg-day)
BAF 109
Import and repackaging
5.5
2.17E-02
1.08E-03
1.73E-02
8.56E-04
Incorporation into paints and coatings - 1-part
reactive coatings
15.4
6.03E-02
2.99E-03
4.79E-02
2.38E-03
Incorporation into paints and coatings - 2-part
reactive coatings
14.0
5.47E-02
2.71E-03
4.35E-02
2.16E-03
Use in paints and coatings at job sites
13.1
5.11E-02
2.53E-03
4.06E-02
2.01E-03
Formulation of TCEP containing reactive
resin
2.1
8.26E-03
4.10E-04
6.56E-03
3.26E-04
Laboratory chemicals
77.1
3.01E-01
1.49E-02
2.40E-01
1.19E-02
" Surface water concentrations based on 90th percentile flow of harmonic mean flow conditions.
Table Apx 1-13. Adult Tribal Fish Ingestion Doses by Scenario Based on a Production Volume of
2,500 lb/year, High-End Release Distribution, Modeled Surface Water Concentrations Based on
90th Percentile Flow, and Two Fish Ingestion Rates
Scenario Name
SWC"
(fig/L)
ADD, ADR
(mg/kg-
day)
BAF 2,198
ADD, ADR
(mg/kg-
day)
BAF 109
LADD
(mg/kg-
day)
BAF 2,198
LADD
(mg/kg-
day)
BAF 109
Current mean fish ingestion rate reported by the Suquamish Tribe (216 g/day)
Import and repackaging
5.5
3.29E-02
1.63E-03
2.62E-02
1.30E-03
Incorporation into paints and coatings - 1-part
reactive coatings
15.4
9.15E-02
4.54E-03
7.27E-02
3.61E-03
Incorporation into paints and coatings - 2-part
reactive coatings
14.0
8.30E-02
4.11E-03
6.59E-02
3.27E-03
Use in paints and coatings at job sites
13.1
7.75E-02
3.84E-03
6.16E-02
3.06E-03
Formulation of TCEP containing reactive resin
2.1
1.25E-02
6.21E-04
9.96E-03
4.94E-04
Laboratory chemicals
77.1
4.57E-01
2.27E-02
3.63E-01
1.80E-02
Heritage fish ingestion rate (1,646 g/day)
Import and repackaging
5.5
2.51E-01
1.24E-02
2.00E-01
9.89E-03
Incorporation into paints and coatings - 1-part
reactive coatings
15.4
6.97E-01
3.46E-02
5.54E-01
2.75E-02
Incorporation into paints and coatings - 2-part
reactive coatings
14.0
6.32E-01
3.14E-02
5.03E-01
2.49E-02
Use in paints and coatings at job sites
13.1
5.91E-01
2.93E-02
4.70E-01
2.33E-02
Formulation of TCEP containing reactive resin
2.1
9.55E-02
4.73E-03
7.59E-02
3.76E-03
Page 548 of 638
-------
Scenario Name
SWC"
(fig/L)
ADD, ADR
(mg/kg-
day)
BAF 2,198
ADD, ADR
(mg/kg-
day)
BAF 109
LADD
(mg/kg-
day)
BAF 2,198
LADD
(mg/kg-
day)
BAF 109
Laboratory chemicals
77.1
3.49
1.73E-01
2.77
1.37E-01
" Surface water concentrations based on 90th percentile flow of harmonic mean flow conditions.
Page 549 of 638
-------
1.4.2 Risk Estimates
Table Apx 1-14. Acute Fish Ingestion Non-cancer Risk Summary Based on 90th Percentile Flow of Harmonic Mean
cou
OES
Acute Oral Non-cancer MOEs
UFs = 30
Life Cycle
Stage/Category
Subcategory
General Population
Subsistence Fishers
Tribes
(Current IR)"
Tribes
(Heritage IR)*
BAF
2,198
BAF
109
BAF
2,198
BAF
109
BAF
2,198
BAF
109
BAF
2,198
BAF
109
Manufacturing/
Import
Import
Repackaging
2,800
56,460
436
8,786
287
5,792
38
760
Processing/
Processing -
Incorporation into
formulation,
mixture, or
reaction product
Flame retardant in: paint
and coating manufacturing
Incorporation into
paints and coatings -
1-part coatings
1,008
20,321
157
3,162
103
2,085
14
274
Incorporation into
paints and coatings -
2-part reactive coatings
1,112
22,414
173
3,488
114
2,300
15
302
Polymers used in aerospace
equipment and products
Formulation of TCEP
containing reactive
resin
1,189
23,986
185
3733
122
2,461
16
323
Commercial use
Laboratory chemicals
Use of laboratory
chemicals
7,361
148,441
1,146
23,100
755
15,229
99
1,998
Paints and coatings
Use of paints and
coatings at job sites
202
4,066
31
633
21
417
3
55
" Current fish consumption rate at 216 g/day based on survey of Suquamish Indian Tribe in Washington (see Section 5.1.3.4.4).
h Heritage fish consumption rate at 1,646 g/day based on study of Kootenai Tribe in Idaho (see Section 5.1.3.4.4).
Page 550 of 638
-------
Table Apx 1-15. Chronic Fish Ingestion Non-cancer Risk Summary Based on 90th Percentile Flow of Harmonic Mean
cou
Gen Pop
Subsistence Fishers*
Tribes (Current)'
Tribes (Heritage)''
Life Cycle
Subcategory
OES
BAF 2,198"
BAF 109"
BAF
BAF
BAF
BAF
BAF
BAF
Stage/Category
CTe
HE
CTe
HE
2,198
109
2,198
109
2,198
109
Manufacturing/
Import
Repackaging
3,553
808
71,639
16,293
126
2,536
83
1,672
11
219
Import
Incorporation
1,279
291
25,784
5,864
45
913
30
602
4
79
Flame
retardant in:
paint and
coating
manufacturing
into paints and
Processing/
Processing -
coatings - 1-part
coatings
Incorporation
into
Formulation,
Incorporation
into paints and
coatings - 2-part
1,410
321
28,440
6,468
50
1,007
33
664
4
87
Mixture, or
reactive coatings
Reaction
Polymers used
Formulation of
9,340
2,124
188,349
42,838
331
6,666
218
4,395
29
577
Product
in aerospace
equipment and
products
TCEP containing
reactive resin
Aerospace
Processing into
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Processing/
Processing -
Incorporation
into Article
equipment and
products and
automotive
articles and
replacement
parts
containing
TCEP
2-part resin
article
Commercial
Use
Laboratory
chemicals
Use of laboratory
chemicals
256
58
5,160
1,173
9
183
6
120
1
16
Paints and
Use of paints and
1,509
343
30,435
6,922
53
1,077
35
710
5
93
coatings
coatings at job
sites
" GP exposure estimates based on general population fish ingestion rate of 22.2 g/day.
b SF exposure estimates based on subsistence fisher ingestion rate of 142.2 g/day.
c Current fish consumption rate at 216 g/day based on survey of Suquamish Indian Tribe in Washington (see Section 5.1.3.4.4).
J Heritage fish consumption rate at 1,646 g/day based on study of Kootenai Tribe in Idaho (see Section 5.1.3.4.4).
e Exposure estimates based on a general population mean fish ingestion rate of 5.04 g/day.
Page 551 of 638
-------
Table Apx 1-16. Lifetime Cancer Risk Summary for Fish Consumption Based on 90th Percentile Flow of Harmonic Mean
cou
OES
Lifetime Cancer Oral Risk Estimates
Life Cycle
Stage/Category
Subcategory
Adult Fish Ingestion General
Population"
Adult Subsistence
Fisher
Tribes
(Current IR)
Tribes
(Heritage IR)
BAF 2,198
BAF 109
BAF
2,198
BAF
109
BAF
2,198
BAF
109
BAF
2,198
BAF
109
CT*
HE
CT*
HE
Manufacturing/
Import
Import
Repackaging
1.50E-05
6.58E-05
7.42E-07
3.26E-06
4.23E-04
2.10E-05
6.41E-04
3.18E-05
4.89E-03
2.42E-04
Processing/
Processing -
Incorporation
into
Formulation,
Mixture, or
Reaction
Product
Flame retardant
in: paint and
coating
manufacturing
Incorporation into
paints and coatings
- 1-part coatings
4.16E-05
1.83E-04
2.06E-06
9.07E-06
1.17E-03
5.83E-05
1.78E-03
8.84E-05
1.36E-02
6.74E-04
Incorporation into
paints and coatings
- 2-part reactive
coatings
3.77E-05
1.66E-04
1.87E-06
8.22E-06
1.07E-03
5.28E-05
1.62E-03
8.01E-05
1.23E-02
6.11E-04
Polymers used in
aerospace
equipment and
products
Formulation of
TCEP containing
reactive resin
5.69E-06
2.50E-05
2.82E-07
1.24E-06
1.61E-04
7.98E-06
2.441- 04
1.21E-05
1.86E-03
9.22E-05
Commercial
Use
Laboratory
chemicals
Use of laboratory
chemicals
2.08E-04
9.14E-04
1.03E-05
4.53E-05
5.87E-03
2.91E-04
8.90E-03
4.42E-04
6.79E-02
3.37E-03
Paints and
coatings
Use of paints and
coatings at job
sites
3.52E-05
1.55E-04
1.75E-06
7.68E-06
9.95E-04
4.94E-05
1.51E-03
7.49E-05
1.15E-02
5.71E-04
" Cancer risk estimates for the adult general population are based on the high-end fish ingestion rate of 22.2 g/day.
b Exposure estimates are based on a general population mean fish ingestion rate of 5.04 g/day.
Page 552 of 638
-------
1.5 Human Milk Pathway
TCEP is predicted to passively accumulate in human milk because it has a small mass (285.48 Da), is
slightly lipophilic (Log P = 1.78), and is a weak base (thus, less likely to be ionized or protein bound).
The key chemical characteristics of TCEP are shown below in Table Apx 1-17. Furthermore,
biomonitoring data confirmed TCEP's presence in human milk (He et al.. 2018a; Kim et al.. 2014;
Sundkvist et al.. 2010). Because of TCEP's potential to transfer to human milk and infants'
susceptibility to its health effects, a quantitative analysis of the milk pathway is necessary to predict
potential risks to infants. TCEP concentrations in milk were estimated based on the maternal doses using
a multi-compartment physiologically based pharmacokinetic (PBPK) model identified by EPA as the
best available model (Verner et al.. 2009; Verner et al.. 2008). hereafter referred to as the Verner Model.
Table Apx 1-17. Key
Chemical Characteristics of TCEP
Key Question or
Decision
Result
Chemical Property or
Population
Current Value Used
for Analysis
Reference(s)
Is the chemical
lipophilic (log P > 1) and
less than 800 Da?
Yes
Average mass
285.49 Da
CompTox Dashboard
(epa.aov) 1 Tris(2-
chloroethyl) phosphate
Log Kow (Log P) from
Scoping review (Measured)
1.78
U.S. EPA (2020b)
Log Kow (Log P) from
other EPA sources
1.44, 1.78, 0.54-1.4
EPA, personal
communication
Log Kow (Log P, Predicted)
1.44108
CompTox Dashboard
(epa.gov) 1 Tris(2-
chloroethvl) phosphate
Is the chemical
hydrophilic and less than
200 Da?
No
Average mass
285.49 Da
CompTox Dashboard
(epa.aov) 1 Tris(2-
chloroethvl) phosphate
Water solubility (measured)
7,820 mg/L at 20 °C
U.S. EPA (2020b)
Is the chemical a weak
base?
Neutral
pH
according
to
reference
pKa
TCEP does not have a
pKa because it does
not have any ionizable
groups. It may have a
pKb, but those are
rarely reported.
PubChem (nih.eov) 1
compound/8295
Phosphorus esters
hydrolysis rates available
NR
U.S. EPA (2020b)
Passive Diffusion
Prediction
Yes
Also supported by
topological polar surface
area (calculated)11
44.8 A
PubChem (nih.eov) 1
compound/8295
Is there evidence of
passive diffusion in
peer-reviewed literature?
No
N/A
NR
N/A
Active Transport
Prediction
No
N/A
NR
N/A
Is there evidence of
active transport?
No
N/A
NR
N/A
Has the chemical been
detected in human milk?
Yes
United States
Range: ND to 0.8
ng/mL
Ma et al. (2019)
Page 553 of 638
-------
Key Question or
Decision
Result
Chemical Property or
Population
Current Value Used
for Analysis
Reference(s)
Women in Australia, Japan,
Philippines, Vietnam, and
Sweden
Range: ND to 0.47
ng/mL
He et al. (2018a)
Central tendency: 0.14
ng/g to 42 ng/g lw
Kim et al. (2014)
Central tendency: 4.9
ng/g lw
Sundkvist et al. (2010)
Is there a measured
value for human milk
partition coefficient?
No
N/A
NR
N/A
" The topological polar surface area of a molecule is defined as the surface sum over all polar atoms in a molecule.
Membrane Dcrmcabilitv is tvmcallv limited when oolar surface area (PSA) exceeds 140 A2. (Matsson and Kihlbere. 2017).
NR - not reported.
1.5.1 Verner Model
The solubility of TCEP in the water of tissue and blood must be considered because it is slightly
lipophilic (log P = 1.78). EPA identified the Verner Model, a multi-compartment PBPK model that
distributes a chemical between different tissue compartments, as appropriate for evaluating infant
exposure to less lipophilic chemicals like TCEP. The Verner Model accounts for every female lifestage
and includes data on maternal height, weight, and age. It also integrates several concurrent physiologic
events that are relevant to infant exposure from milk (e.g., pre- and postpartum changes in maternal
physiology, lactation, infant growth) and inputs physiological parameters, including organ volume,
composition, and blood flow throughout a woman's entire life. Note that the Verner Model was
validated using only data on persistent organic pollutants levels measured in mothers and infants from a
Northern Quebec Inuit population (Verner et al.. 2009). It was not validated using data on TCEP, which
were not available.
The Verner model describes the period from the beginning of the mother's life to the first year of the
infant's life. As shown in Figure Apx 1-6, the model consists of a total of 14 compartments: 9 maternal
(uterus, brain, richly perfused tissue, poorly perfused tissue, adipose tissue, mammary tissue, liver,
placenta, and fetus) and 5 infantile (brain, richly perfused tissue, poorly perfused tissue, adipose tissue,
and liver). Distribution of the chemical is driven by blood flow and the partitioning between the blood
and the tissues.
Page 554 of 638
-------
Mother
Initial Body Burden
/ ,
Richly Perfused
A
/
1
Poorly Perfused
|
r
i
I
Adipose Tissue
k-
,
Human Milk
Metabolism
FigureApx 1-6. Compartments and Exposure Routes for Verner Model
Adapted from (Verner et al.. 2009).
EPA implemented the Verner Model in the R programming language to enable running the model using
modern R packages. The model was written as three systems of ordinary differential equations (ODEs),
corresponding to preconception, pregnancy, and breastfeeding. The number of compartments included in
preconception, pregnancy, and breastfeeding are 7, 9, and 12, respectively. In addition, the following
additional updates were introduced into the R code:
Discontinuities related to physiological terms at ages 3 and 18 were corrected.
Mass balance tables were introduced for quality assurance evaluation.
Brain volume parameters were added (personal communication) (Verner et al.. 2008).
A batch version of the code was developed to run several exposure scenarios consecutively.
Graphics were elaborated to visualize three key stages: conception, birth, and lactation.
Milk intake rates updated using EPA's Exposure Factors Handbook (U.S. EPA. 201 la).
Model output expanded to include daily infant dose.
Model computes peak and average infant dose for each age group within the first year of life.
The model inputs are shown in Table Apx 1-18 below.
Page 555 of 638
-------
Table Apx 1-18. Data Input Requirements for the Multi-compartment Model
Input
Organs or Data
Data Source(s)
Blood flow
Mother: fetus, placenta, uterus, brain, richly
perfused tissue, poorly perfused tissue,
adipose tissue, mammary tissue, liver, heart
Infant: brain, richly perfused tissue, poorly
perfused tissue, adipose tissue, liver, heart
Calculated from equations in (Verner et al..
2009; Verner et al.. 2008); blood flow to brain
was not published and estimated based on
correspondences with author
Organ volume
Mother: fetus, placenta, uterus, brain, richly
perfused tissue, poorly perfused tissue,
adipose tissue, mammary tissue, liver
Infant: brain, richly perfused tissue, poorly
perfused tissue, adipose tissue, liver
Calculated from equations in (Verner et al..
2009; Verner et al.. 2008). Changes made to
skeletal muscles (part of poorly perfused
tissue) and extra fat, mammary, and uterine
volume at end of pregnancy to keep
parameters continuous
Fraction of lipid or
water in tissue
Mother: blood, brain, liver, adipose tissue,
richly perfused tissue, poorly perfused tissue,
mammary tissue, uterus, placenta
Infant: blood, adipose tissue, liver, richly
perfused tissue, poorly perfused tissue, brain
(Verner et al.. 2009; Verner et al.. 2008; Price
etal.. 2003; White etal.. 1991)
Tissue :blood
partition
coefficients
Mother: fetus, placenta, uterus, brain, richly
perfused tissue, poorly perfused tissue,
adipose tissue, mammary tissue, liver
Infant: brain, richly perfused tissue, poorly
perfused tissue, adipose tissue, liver
Calculated from Kow, fraction of lipid or
water in tissue of interest, and equation in
(Verner et al.. 2008)
Milk:blood partition
coefficient
Same formula used for tissue :blood
coefficients
Calculated from Kow, fraction of lipid or
water in milk, and equations in (Verner et al..
2008)
Fraction of lipids in
milk
Function of number of days post-partum, or
age of the child
(Verner et al.. 2008)
Half-life (TCEP)
17.64 hours
Half-life is used to calculate a hepatic
extraction ratio that varies by age because it
considers blood and tissue volumes that
change by age.
Half-life value estimated from a one-
compartment model
https://comptox.epa.eov/dashboard/cIiemical/
adme-ivive-subtab/DTXSID5021411
Oral dose
Default/User input
Derived from occupational, consumer, and
general population doses adjusted for body
weight representative of women of
reproductive age
Duration of
breastfeeding
Default/user input
One year is the default.
Volume of
breastfeeding
Default/user input
(Verner et al.. 2009)
Description of Absorption, Distribution, and Excretion Parameters
The model is composed of three different stages: pre-conception, pregnancy, and breastfeeding. Each
dA,
model solves the rate of change of the amount —— of the chemical in compartment t (tissue) as listed in,
Page 556 of 638
-------
Table Apx 1-19 where At denotes the amount of chemical in the tissue. These rates of change are given
in terms of the blood flow to the tissue Qt, the compartment concentration Ct, the tissue:blood partition
coefficient Pt.b, and the arterial blood concentration Ca, as collectively defined under EquationApx 1-8
below. The distribution of the chemical can be described by mass balance equations for tissue t as
described in Verner et al. (2008) as
EquationApx 1-8.
with this sum being taken over all tissues. Here, Qc denotes the cardiac blood flow and Cvt denotes the
tissue venous blood concentration. The tissue:blood partition coefficients can be computed according to
Verner et al. (2008) by
where Kow denotes the octanol-water partition coefficient of the chemical under consideration, Flt and
Fwt denote the time-varying percentages of lipid and water, respectively, in compartment t. Flb and
Fwb denote the percentages of lipid and water, respectively, in blood.
The mass balance equation for the liver compartment has a slightly different form, as it has an
absorption and metabolism term. It is given by Verner et al. (2008) as
where Qi is the blood flow to the liver and RAM represents the metabolism in |ig/day. To compute this,
the volume of distribution is first calculated.
where Vbi00d denotes the volume of blood in the mother, computed according to the Nadler equation
(Sharma and Sharma. 2023). rp is richly perfused, pp is poorly perfused, u is uterus,/is fat, / is liver,
and mam is mammary tissue. This is used to compute additional parameters defined in (Verner et al..
2008). The clearance is
where HL denotes the half-life of the chemical in days. This is used to compute the quantity Ehage as
The arterial blood concentration is computed as
= Kow ¦ Flt + Fwt
t]b ~ Kow ¦ Flb + Fwb
Page 557 of 638
-------
_ CLage
C'Lage ~ qi -
which in turn is used to compute the intrinsic clearance value
CLintc=--
VI \1 Ehage)
From here, the hepatic extraction is computed by
CLintr-Vl
Lj (X —
CLintc'Vl+Ql5
which is used to compute the metabolism rate measured in |ig/day.
RAM = Ql • Eh • Ca,
To solve this system of differential equations, organ volumes and blood flows are required for all times.
The system is solved numerically using the ODE function in the deSolve package in R. The output of
the model is a chemical amount and concentration in each organ compartment, as well as the TCEP
concentrations in milk for the entire time period of the simulation.
1.5.2 Milk Ingestion Rates by Age
Milk ingestion rates by age are provided in Table 15-1 of the Exposure Factors Handbook (U.S. EPA.
2011a) and presented in Table Apx 1-19.
Table Apx 1-19. Mean and Upper Milk Ingestion Rates by Age
Age Group
Milk Ingestion (mL/kg day)
Mean
Upper (95th percentile)
Birth to <1 month
150
220
1 to <3 month
140
190
3 to <6 month
110
150
6 to <12 month
83
130
Birth to <1 year
104.8
152.5
1.5.3 Modeled TCEP Concentrations in Milk
Four biomonitoring studies demonstrated the presence of TCEP in human milk. One U.S. study
measured a mean wet weight concentration of 0.036 ng/mL among 100 samples, with a range from 0-
0.8 ng/mL (Ma et al.. 2019). Ma et al. (2019) set non-detects to 0 and concentrations below the limit of
quantification (LOQ) to half the value of the LOQ. Among non-U. S. studies, only one measured wet
weight concentrations from three milk samples collected in Australia, and concentrations ranged from
non-detect (<0.13 ng/mL) to 0.47 ng/mL (He et al.. 2018a). He et al. (2018a) assigned half the value of
the method detection limit for all non-detects. Two other non-US studies measured lipid weight
concentrations that ranged from non-detect to 512 ng/g (average 0.14-42 ng/g) in (Kim et al.. 2014) and
2.1 to 8.2 ng/g (median 4.9 ng/g) in (Sundkvist et al.. 2010). These two studies' treatment of non-detects
are not discussed because they report lipid weight concentrations that cannot be compared to the outputs
from the Verner Model. The Verner Model estimates wet weight concentrations, thus modeled
Page 558 of 638
-------
concentrations can only be compared with measured concentrations by (Ma et al.. 2019) and (He et al..
2018a). The range of the wet weight concentrations across each COU/OES for each maternal group is
presented in TableApx 1-17. In general, the lower and upper bound of the modeled concentrations are
three orders of magnitudes below and five orders of magnitudes above measured concentrations,
respectively.
Table Apx 1-20. Comparison of the Range of Measured and Modeled TCEP Concentrations in
Human Milk
Maternal Group
Modeled Concentrations (mg/mL)
Measured Concentrations (mg/mL)
Consumer
3.96E-08 to 2.62E-04
0 to 8E-07 (based on 100 samples
collected in U.S.)
<1.3E-07 to 4.7E-07 (based on three
samples collected in Australia)
Occupational
1.96E-10 to 1.13E-03
General population
1.83E-10 to 5.22E-04
Tribal populations
1.79E-06 to 2.93E-02
1.5.4 Infant Exposure Estimate
Page 559 of 638
-------
Table Apx 1-21. Average Infant Doses via I
unian Milk Exposure from
Maternal Consumer Use Scenarios
COU Subcategory and Consumer
Exposure Scenarios
Maternal Dose
(mg/kg-day)"b
Milk Intake
Rate Type
Birth to <1 Month
(mg/kg-day)
1 to <3 Month
(mg/kg-day)
3 to <6 Month
(mg/kg-day)
6 to 12 Month
(mg/kg-day)
Birth to 12 Month
(mg/kg-day)
Fabric textile, leather products not
covered elsewhere (carpet back
coating)
6.12E-03
Mean
1.01E-04
1.03E-04
9.23E-05
8.39E-05
9.06E-05
Fabric textile, leather products not
covered elsewhere (textile for
children's play structures)
9.02E-02
Mean
1.49E-03
1.52E-03
1.36E-03
1.24E-03
1.33E-03
Building/construction materials not
covered elsewhere (roofing
insulation)
1.74
Mean
2.87E-02
2.92E-02
2.62E-02
2.38E-02
2.57E-02
Building/construction materials not
covered elsewhere (acoustic ceiling)
1.40E-01
Mean
2.31E-03
2.35E-03
2.10E-03
1.91E-03
2.07E-03
Foam seating and bedding product
(foam automobile)
6.85E-03
Mean
1.13E-04
1.15E-04
1.03E-04
9.39E-05
1.01E-04
Foam seating and bedding product
(foam living room)
1.53E-02
Mean
2.53E-04
2.58E-04
2.31E-04
2.10E-04
2.27E-04
Foam seating and bedding product
(mattress)
1.95E-03
Mean
3.22E-05
3.27E-05
2.93E-05
2.67E-05
2.88E-05
Foam seating and bedding product
(foam - other - toy block)
1.06E-03
Mean
1.75E-05
1.78E-05
1.59E-05
1.45E-05
1.56E-05
Building/construction materials -
wood and engineered wood products
(wood flooring)
1.51
Mean
2.50E-02
2.55E-02
2.28E-02
2.07E-02
2.24E-02
Building/construction materials -
wood and engineered wood products
(wooden TV stand)
1.01E-01
Mean
1.68E-03
1.70E-03
1.53E-03
1.39E-03
1.50E-03
Fabric textile, leather products not
covered elsewhere (carpet back
coating)
6.12E-03
Upper
1.48E-04
1.39E-04
1.25E-04
1.30E-04
1.32E-04
Fabric textile, leather products not
covered elsewhere (textile for
children's play structures)
9.02E-02
Upper
2.18E-03
2.05E-03
1.85E-03
1.92E-03
1.95E-03
Building/construction materials not
covered elsewhere (roofing
insulation)
1.74
Upper
4.20E-02
3.95E-02
3.56E-02
3.70E-02
3.75E-02
Building/construction materials not
covered elsewhere (acoustic ceiling)
1.40E-01
Upper
3.38E-03
3.18E-03
2.86E-03
2.98E-03
3.02E-03
Page 560 of 638
-------
COU Subcategory and Consumer
Exposure Scenarios
Maternal Dose
(mg/kg-day)"b
Milk Intake
Rate Type
Birth to <1 Month
(mg/kg-day)
1 to <3 Month
(mg/kg-day)
3 to <6 Month
(mg/kg-day)
6 to 12 Month
(mg/kg-day)
Birth to 12 Month
(mg/kg-day)
Foam seating and bedding product
(foam automobile)
6.85E-03
Upper
1.66E-04
1.56E-04
1.40E-04
1.46E-04
1.48E-04
Foam seating and bedding product
(foam living room)
1.53E-02
Upper
3.71E-04
3.49E-04
3.14E-04
3.27E-04
3.31E-04
Foam seating and bedding product
(mattress)
1.95E-03
Upper
4.71E-05
4.43E-05
3.99E-05
4.15E-05
4.20E-05
Foam seating and bedding product
(foam - other - toy block)
1.06E-03
Upper
2.56E-05
2.41E-05
2.17E-05
2.25E-05
2.28E-05
Building/construction materials -
wood and engineered wood products
(wood flooring)
1.51
Upper
3.67E-02
3.45E-02
3.10E-02
3.23E-02
3.27E-02
Building/construction materials -
wood and engineered wood products
(wooden TV stand)
1.01E-01
Upper
2.45E-03
2.31E-03
2.07E-03
2.16E-03
2.19E-03
" Consumer maternal doses were combined across oral, dermal, and inhalation routes. For inhalation, CEM 3.2 already calculates a dose in mg/kg-day, as shown in
Section 5.1.2.3 for consumers.
h Chronic maternal doses are the most relevant durations for building and construction materials, fabric and textile products, and foam seating and bedding products
because they are typically used over a longer time frame than other types of consumer products with direct applications (e.g., household cleaners, solvents).
Page 561 of 638
-------
Table Apx 1-22. Average Infant Doses from Maternal Workers Based on Mean Milk Intake Rate
OES
Route
Maternal
Exposure
Duration
Maternal Dose
(mg/kg-day)
Birth to <1
Month
(mg/kg-day)
1 to <3
Month
(mg/kg-day)
3 to <6
Month
(mg/kg-day)
6 to 12
Month
(mg/kg-day)
Birth to 12
Month
(mg/kg-day)
Import and repackaging
Dermal,
Inhalation
(High-end)
Chronic
1.57E-01
2.59E-03
2.63E-03
2.36E-03
2.15E-03
2.32E-03
Incorporation into paints and coatings - 1-part
coatings
Chronic
8.38E-01
1.39E-02
1.41E-02
1.26E-02
1.15E-02
1.24E-02
Incorporation into paints and coatings - 2-part
reactive coatings
Chronic
8.53E-02
1.41E-03
1.43E-03
1.29E-03
1.17E-03
1.26E-03
Processing - Formulation of TCEP into 2-part
reactive resins
Chronic
1.73E-01
2.86E-03
2.90E-03
2.60E-03
2.37E-03
2.56E-03
Processing - Processing into 2-part resin article
Chronic
2.18
3.60E-02
3.66E-02
3.28E-02
2.98E-02
3.22E-02
Processing - Recycling electronics
Chronic
1.37E-04
2.26E-06
2.30E-06
2.06E-06
1.87E-06
2.03E-06
Commercial use - Paints and coatings - Spray
(1-part, 250-day application)
Chronic
1.45
2.40E-02
2.44E-02
2.18E-02
1.99E-02
2.14E-02
Commercial use - Paints and coatings - Spray
(2-part reactive, 250-day application)
Chronic
7.25
1.20E-01
1.22E-01
1.09E-01
9.93E-02
1.07E-01
Laboratory chemicals
Chronic
4.35
7.20E-02
7.32E-02
6.56E-02
5.96E-02
6.44E-02
Industrial/Commercial use - Installation of
articles, chronic, inhalation
Inhalation
(High-end)
Chronic
1.35E-06
2.23E-08
2.27E-08
2.04E-08
1.85E-08
2.00E-08
Import and repackaging
Dermal,
Inhalation
(High-end)
Intermediate
1.86
3.07E-02
3.12E-02
2.80E-02
2.55E-02
2.75E-02
Incorporation into paints and coatings - 1-part
coatings
Intermediate
5.84
9.65E-02
9.81E-02
8.79E-02
8.00E-02
8.64E-02
Incorporation into paints and coatings - 2-part
reactive coatings
Intermediate
5.74E-01
9.50E-03
9.65E-03
8.65E-03
7.87E-03
8.50E-03
Processing - Formulation of TCEP into 2-part
reactive resins
Intermediate
1.63
2.70E-02
2.75E-02
2.46E-02
2.24E-02
2.42E-02
Processing - Processing into 2-part resin article
Intermediate
2.33
3.86E-02
3.92E-02
3.51E-02
3.19E-02
3.45E-02
Processing - Recycling electronics
Intermediate
1.47E-04
2.42E-06
2.46E-06
2.21E-06
2.01E-06
2.17E-06
Commercial use - Paints and coatings - Spray
(1-part, 250-day application)
Intermediate
1.55
2.57E-02
2.61E-02
2.34E-02
2.13E-02
2.30E-02
Commercial use - Paints and coatings - Spray
(2-part reactive, 250-day application)
Intermediate
7.76
1.28E-01
1.30E-01
1.17E-01
1.06E-01
1.15E-01
Laboratory chemicals
Intermediate
5.83
9.63E-02
9.79E-02
8.78E-02
7.98E-02
8.62E-02
Industrial/Commercial use - Installation of
aerospace products, chronic, inhalation
Inhalation
(High-end)
Intermediate
1.45E-06
2.39E-08
2.43E-08
2.18E-08
1.98E-08
2.14E-08
Page 562 of 638
-------
Table Apx 1-23. Average Infant Doses from Maternal Workers Based on IP
jper Milk Intake Rate
OES
Route
Maternal
Exposure
Duration
Maternal Dose
(mg/kg-day)
Birth to <1
Month
(mg/kg-day)
1 to <3
Month
(mg/kg-day)
3 to <6
Month
(mg/kg-day)
6 to 12
Month
(mg/kg-day)
Birth to 12
Month
(mg/kg-day)
Import and repackaging
Dermal,
Inhalation
(High-end)
Chronic
1.57E-01
3.79E-03
3.56E-03
3.21E-03
3.34E-03
3.38E-03
Incorporation into paints and coatings - 1-part
coatings
Chronic
8.38E-01
2.03E-02
1.91E-02
1.72E-02
1.79E-02
1.81E-02
Incorporation into paints and coatings - 2-part
reactive coatings
Chronic
8.53E-02
2.06E-03
1.94E-03
1.75E-03
1.82E-03
1.84E-03
Processing - Formulation of TCEP into 2-part
reactive resins
Chronic
1.73E-01
4.18E-03
3.93E-03
3.54E-03
3.68E-03
3.73E-03
Processing - Processing into 2-part resin article
Chronic
2.18
5.27E-02
4.96E-02
4.46E-02
4.64E-02
4.70E-02
Processing - Recycling electronics
Chronic
1.37E-04
3.31E-06
3.11E-06
2.80E-06
2.92E-06
2.95E-06
Commercial use - Paints and coatings - Spray
(1-part, 250-day application)
Chronic
1.45
3.51E-02
3.30E-02
2.97E-02
3.09E-02
3.13E-02
Commercial use - Paints and coatings - Spray
(2-part reactive, 250-day application)
Chronic
7.25
1.75E-01
1.65E-01
1.48E-01
1.54E-01
1.56E-01
Laboratory chemicals
Chronic
4.35
1.05E-01
9.91E-02
8.92E-02
9.28E-02
9.40E-02
Industrial/Commercial use - Installation of
articles, chronic, inhalation
Inhalation
(High-end)
Chronic
1.35E-06
3.27E-08
3.08E-08
2.77E-08
2.88E-08
2.92E-08
Import and repackaging
Dermal,
Inhalation
(High-end)
Intermediate
1.86
4.50E-02
4.23E-02
3.81E-02
3.96E-02
4.62E-02
Incorporation into paints and coatings - 1-part
coatings
Intermediate
5.84
1.41E-01
1.33E-01
1.20E-01
1.24E-01
1.45E-01
Incorporation into paints and coatings - 2-part
reactive coatings
Intermediate
5.74E-01
1.39E-02
1.31E-02
1.18E-02
1.22E-02
1.43E-02
Processing - Formulation of TCEP into 2-part
reactive resins
Intermediate
1.63
3.95E-02
3.72E-02
3.35E-02
3.48E-02
4.07E-02
Processing - Processing into 2-part resin article
Intermediate
2.33
5.65E-02
5.31E-02
4.78E-02
4.97E-02
5.80E-02
Processing - Recycling electronics
Intermediate
1.47E-04
3.55E-06
3.33E-06
3.00E-06
3.12E-06
3.65E-06
Commercial use - Paints and coatings - Spray
(1-part, 250-day application)
Intermediate
1.55
3.76E-02
3.53E-02
3.18E-02
3.31E-02
3.86E-02
Commercial use - Paints and coatings - Spray
(2-part reactive, 250-day application)
Intermediate
7.76
1.88E-01
1.77E-01
1.59E-01
1.65E-01
1.93E-01
Laboratory chemicals
Intermediate
5.83
1.41E-01
1.33E-01
1.19E-01
1.24E-01
1.45E-01
Industrial/commercial use - Installation of
articles, chronic, inhalation
Inhalation
(High-end)
Intermediate
1.45E-06
3.50E-08
3.29E-08
2.96E-08
3.08E-08
3.60E-08
Page 563 of 638
-------
TableApx 1-24. Average Infant Doses via Human Milk Exposure from Maternal General Population Oral Exposures Based on Mean
Milk Intake Rate
COUs/OES
Route
Maternal Dose
(mg/kg-day)
Birth to <1
Month
(mg/kg-day)
1 to <3
Month
(mg/kg-day)
3 to <6
Month
(mg/kg-day)
6 to 12
Month
(mg/kg-day)
Birth to 12
Month
(mg/kg-day)
Import and repackaging
Gen Pop Fish Ingestion,
High B AF
6.37E-01
1.05E-02
1.07E-02
9.60E-03
8.73E-03
9.43E-03
Incorporation into paints and coatings - 1-part
coatings
Gen Pop Fish Ingestion,
High B AF
2.82
4.66E-02
4.74E-02
4.25E-02
3.86E-02
4.17E-02
Incorporation into paints and coatings - 2-part
reactive coatings
Gen Pop Fish Ingestion,
High B AF
2.56
4.23E-02
4.30E-02
3.86E-02
3.51E-02
3.79E-02
Use in paints and coatings at job sites
Gen Pop Fish Ingestion,
High B AF
1.50
2.48E-02
2.52E-02
2.26E-02
2.05E-02
2.22E-02
Formulation of TCEP containing reactive resin
Gen Pop Fish Ingestion,
High B AF
3.58
5.92E-02
6.02E-02
5.39E-02
4.90E-02
5.30E-02
Laboratory chemicals
Gen Pop Fish Ingestion,
High B AF
2.55E-02
4.22E-04
4.29E-04
3.84E-04
3.49E-04
3.77E-04
Import and repackaging
Gen Pop Fish Ingestion,
Low BAF
3.16E-02
5.23E-04
5.31E-04
4.76E-04
4.33E-04
4.68E-04
Incorporation into paints and coatings - 1-part
coatings
Gen Pop Fish Ingestion,
Low BAF
1.40E-01
2.32E-03
2.35E-03
2.11E-03
1.92E-03
2.07E-03
Incorporation into paints and coatings - 2-part
reactive coatings
Gen Pop Fish Ingestion,
Low BAF
1.27E-01
2.10E-03
2.13E-03
1.91E-03
1.74E-03
1.88E-03
Use in paints and coatings at job sites
Gen Pop Fish Ingestion,
Low BAF
7.44E-02
1.23E-03
1.25E-03
1.12E-03
1.02E-03
1.10E-03
Formulation of TCEP containing reactive resin
Gen Pop Fish Ingestion,
Low BAF
1.78E-01
2.94E-03
2.99E-03
2.68E-03
2.44E-03
2.63E-03
Laboratory chemicals
Gen Pop Fish Ingestion,
Low BAF
1.27E-03
2.10E-05
2.13E-05
1.91E-05
1.74E-05
1.88E-05
Import and repackaging
Undiluted Drinking Water
3.16E-05
5.23E-07
5.31E-07
4.76E-07
4.33E-07
4.68E-07
Incorporation into paints and coatings - 1-part
coatings
Undiluted Drinking Water
1.40E-04
2.31E-06
2.35E-06
2.11E-06
1.92E-06
2.07E-06
Incorporation into paints and coatings - 2-part
reactive coatings
Undiluted Drinking Water
1.26E-04
2.08E-06
2.12E-06
1.90E-06
1.73E-06
1.86E-06
Use in paints and coatings at job sites
Undiluted Drinking Water
7.42E-05
1.23E-06
1.25E-06
1.12E-06
1.02E-06
1.10E-06
Formulation of TCEP containing reactive resin
Undiluted Drinking Water
1.77E-04
2.93E-06
2.97E-06
2.67E-06
2.42E-06
2.62E-06
Laboratory chemicals
Undiluted Drinking Water
1.26E-06
2.08E-08
2.12E-08
1.90E-08
1.73E-08
1.86E-08
Page 564 of 638
-------
TableApx 1-25. Average Infant Doses via Human Milk Exposure from Maternal General Population Oral Exposures Based on
Upper Milk Intake Rate
COUs/OES
Route
Maternal Dose
(mg/kg-day)
Birth to <1
Month
(mg/kg-day)
1 to <3
Month
(mg/kg-day)
3 to <6
Month
(mg/kg-day)
6 to 12
Month
(mg/kg-day)
Birth to 12
Month
(mg/kg-day)
Import and repackaging
Gen Pop Fish Ingestion,
High B AF
6.37E-01
1.54E-02
1.45E-02
1.30E-02
1.36E-02
1.38E-02
Incorporation into paints and coatings - 1-part
coatings
Gen Pop Fish Ingestion,
High B AF
2.82
6.83E-02
6.42E-02
5.78E-02
6.01E-02
6.09E-02
Incorporation into paints and coatings - 2-part
reactive coatings
Gen Pop Fish Ingestion,
High B AF
2.56
6.20E-02
5.83E-02
5.24E-02
5.45E-02
5.53E-02
Use in paints and coatings at job sites
Gen Pop Fish Ingestion,
High B AF
1.50
3.63E-02
3.41E-02
3.07E-02
3.20E-02
3.24E-02
Formulation of TCEP containing reactive resin
Gen Pop Fish Ingestion,
High B AF
3.58
8.66E-02
8.15E-02
7.33E-02
7.63E-02
7.73E-02
Laboratory chemicals
Gen Pop Fish Ingestion,
High B AF
2.55E-02
6.17E-04
5.80E-04
5.22E-04
5.43E-04
5.51E-04
Import and repackaging
Gen Pop Fish Ingestion,
Low BAF
3.16E-02
7.65E-04
7.19E-04
6.47E-04
6.73E-04
6.82E-04
Incorporation into paints and coatings - 1-part
coatings
Gen Pop Fish Ingestion,
Low BAF
1.40E-01
3.39E-03
3.19E-03
2.87E-03
2.98E-03
3.02E-03
Incorporation into paints and coatings - 2-part
reactive coatings
Gen Pop Fish Ingestion,
Low BAF
1.27E-01
3.07E-03
2.89E-03
2.60E-03
2.71E-03
2.74E-03
Use in paints and coatings at job sites
Gen Pop Fish Ingestion,
Low BAF
7.44E-02
1.80E-03
1.69E-03
1.52E-03
1.59E-03
1.61E-03
Formulation of TCEP containing reactive resin
Gen Pop Fish Ingestion,
Low BAF
1.78E-01
4.31E-03
4.05E-03
3.65E-03
3.79E-03
3.84E-03
Laboratory chemicals
Gen Pop Fish Ingestion,
Low BAF
1.27E-03
3.07E-05
2.89E-05
2.60E-05
2.71E-05
2.74E-05
Import and repackaging
Undiluted Drinking Water
3.16E-05
7.65E-07
7.19E-07
6.47E-07
6.73E-07
6.82E-07
Incorporation into paints and coatings -1-part
coatings
Undiluted Drinking Water
1.40E-04
3.39E-06
3.19E-06
2.87E-06
2.98E-06
3.02E-06
Incorporation into paints and coatings - 2-part
reactive coatings
Undiluted Drinking Water
1.26E-04
3.05E-06
2.87E-06
2.58E-06
2.68E-06
2.72E-06
Use in paints and coatings at job sites
Undiluted Drinking Water
7.42E-05
1.80E-06
1.69E-06
1.52E-06
1.58E-06
1.60E-06
Formulation of TCEP containing reactive resin
Undiluted Drinking Water
1.77E-04
4.28E-06
4.03E-06
3.63E-06
3.77E-06
3.82E-06
Laboratory chemicals
Undiluted Drinking Water
1.26E-06
3.05E-08
2.87E-08
2.58E-08
2.68E-08
2.72E-08
Page 565 of 638
-------
TableApx 1-26. Average Infant Doses via Human Milk Exposure from Maternal Tribal Fish Ingestion Based on Mean Milk Intake
Rate
Maternal Dose
(mg/kg-day)
Birth to <1
1 to <3
3 to <6
6 to 12
Birth to 12
COUs/OES
Route
Month
(mg/kg-day)
Month
(mg/kg-day)
Month
(mg/kg-day)
Month
(mg/kg-day)
Month
(mg/kg-day)
Import and repackaging
Current IR,
HighBAF
6.21
1.03E-01
1.04E-01
9.36E-02
8.51E-02
9.19E-02
Incorporation into paints and coatings - 1-part coatings
Current IR,
HighBAF
2.75E01
4.55E-01
4.62E-01
4.14E-01
3.77E-01
4.07E-01
Incorporation into paints and coatings - 2-part reactive
Current IR,
2.49E01
4.12E-01
4.18E-01
3.75E-01
3.41E-01
3.68E-01
coatings
HighBAF
Use in paints and coatings at job sites
Current IR,
HighBAF
1.46E01
2.41E-01
2.45E-01
2.20E-01
2.00E-01
2.16E-01
Formulation of TCEP containing reactive resin
Current IR,
HighBAF
3.49E01
5.77E-01
5.87E-01
5.26E-01
4.78E-01
5.16E-01
Laboratory chemicals
Current IR,
HighBAF
2.49E-01
4.12E-03
4.18E-03
3.75E-03
3.41E-03
3.68E-03
Import and repackaging
Current IR,
Low BAF
3.08E-01
5.09E-03
5.18E-03
4.64E-03
4.22E-03
4.56E-03
Incorporation into paints and coatings - 1-part coatings
Current IR,
Low BAF
1.36
2.25E-02
2.29E-02
2.05E-02
1.86E-02
2.01E-02
Incorporation into paints and coatings - 2-part reactive
Current IR,
1.24
2.05E-02
2.08E-02
1.87E-02
1.70E-02
1.83E-02
coatings
Low BAF
Use in paints and coatings at job sites
Current IR,
Low BAF
7.25E-01
1.20E-02
1.22E-02
1.09E-02
9.93E-03
1.07E-02
Formulation of TCEP containing reactive resin
Current IR,
Low BAF
1.73
2.86E-02
2.91E-02
2.61E-02
2.37E-02
2.56E-02
Laboratory chemicals
Current IR,
Low BAF
1.23E-02
2.03E-04
2.07E-04
1.85E-04
1.68E-04
1.82E-04
Import and repackaging
Heritage IR,
HighBAF
4.73E01
7.82E-01
7.95E-01
7.13E-01
6.48E-01
7.00E-01
Incorporation into paints and coatings -1-part coatings
Heritage IR,
HighBAF
2.10E02
3.47
3.53
3.16
2.88
3.11
Incorporation into paints and coatings - 2-part reactive
coatings
Heritage IR,
HighBAF
1.91E02
3.16
3.21
2.88
2.62
2.83
Use in paints and coatings at job sites
Heritage IR,
HighBAF
1.11E02
1.84
1.87
1.67
1.52
1.64
Formulation of TCEP containing reactive resin
Heritage IR,
HighBAF
2.66E02
4.40
4.47
4.01
3.64
3.94
Page 566 of 638
-------
COUs/OES
Route
Maternal Dose
(mg/kg-day)
Birth to <1
Month
(mg/kg-day)
1 to <3
Month
(mg/kg-day)
3 to <6
Month
(mg/kg-day)
6 to 12
Month
(mg/kg-day)
Birth to 12
Month
(mg/kg-day)
Laboratory chemicals
Heritage IR,
High B AF
1.89
3.13E-02
3.18E-02
2.85E-02
2.59E-02
2.80E-02
Import and repackaging
Heritage IR,
Low BAF
2.34
3.87E-02
3.93E-02
3.53E-02
3.21E-02
3.46E-02
Incorporation into paints and coatings -1-part coatings
Heritage IR,
Low BAF
1.04E01
1.72E-01
1.75E-01
1.57E-01
1.42E-01
1.54E-01
Incorporation into paints and coatings - 2-part reactive
coatings
Heritage IR,
Low BAF
9.43
1.56E-01
1.58E-01
1.42E-01
1.29E-01
1.40E-01
Use in paints and coatings at job sites
Heritage IR,
Low BAF
5.52
9.13E-02
9.28E-02
8.32E-02
7.56E-02
8.17E-02
Formulation of TCEP containing reactive resin
Heritage IR,
Low BAF
1.32E01
2.18E-01
2.22E-01
1.99E-01
1.81E-01
1.95E-01
Laboratory chemicals
Heritage IR,
Low BAF
9.41E-02
1.56E-03
1.58E-03
1.42E-03
1.29E-03
1.39E-03
TableApx 1-27. Average Infant Doses via Human Milk Exposure from Maternal Tribal Fish Ingestion Based on Upper Milk Intake
Rate
COUs/OES
Route
Maternal Dose
(mg/kg-day)
Birth to <1
Month
(mg/kg-day)
1 to <3
Month
(mg/kg-day)
3 to <6
Month
(mg/kg-day)
6 to 12
Month
(mg/kg-day)
Birth to 12
Month
(mg/kg-day)
Import and repackaging
Current IR, High BAF
6.21
1.50E-01
1.41E-01
1.27E-01
1.32E-01
1.34E-01
Incorporation into paints and coatings - 1-part
coatings
Current IR, High BAF
2.75E01
6.66E-01
6.26E-01
5.63E-01
5.86E-01
5.94E-01
Incorporation into paints and coatings - 2-part
reactive coatings
Current IR, High BAF
2.49E01
6.03E-01
5.67E-01
5.10E-01
5.31E-01
5.38E-01
Use in paints and coatings at job sites
Current IR, High BAF
1.46E01
3.53E-01
3.32E-01
2.99E-01
3.11E-01
3.15E-01
Formulation of TCEP containing reactive resin
Current IR, High BAF
3.49E01
8.45E-01
7.94E-01
7.15E-01
7.44E-01
7.53E-01
Laboratory chemicals
Current IR, High BAF
2.49E-01
6.03E-03
5.67E-03
5.10E-03
5.31E-03
5.38E-03
Import and repackaging
Current IR, Low BAF
3.08E-01
7.45E-03
7.01E-03
6.31E-03
6.56E-03
6.65E-03
Incorporation into paints and coatings - 1-part
coatings
Current IR, Low BAF
1.36
3.29E-02
3.10E-02
2.79E-02
2.90E-02
2.94E-02
Incorporation into paints and coatings - 2-part
reactive coatings
Current IR, Low BAF
1.24
3.00E-02
2.82E-02
2.54E-02
2.64E-02
2.68E-02
Page 567 of 638
-------
COUs/OES
Route
Maternal Dose
(mg/kg-day)
Birth to <1
Month
(mg/kg-day)
1 to <3
Month
(mg/kg-day)
3 to <6
Month
(mg/kg-day)
6 to 12
Month
(mg/kg-day)
Birth to 12
Month
(mg/kg-day)
Use in paints and coatings at job sites
Current IR, Low BAF
7.25E-01
1.75E-02
1.65E-02
1.48E-02
1.54E-02
1.57E-02
Formulation of TCEP containing reactive resin
Current IR, Low BAF
1.73
4.19E-02
3.94E-02
3.54E-02
3.69E-02
3.73E-02
Laboratory chemicals
Current IR, Low BAF
1.23E-02
2.98E-04
2.80E-04
2.52E-04
2.62E-04
2.66E-04
Import and repackaging
Heritage IR, High BAF
4.73E01
1.14
1.08
9.69E-01
1.01
1.02
Incorporation into paints and coatings -1-part
coatings
Heritage IR, High BAF
2.10E02
5.08
4.78
4.30
4.47
4.53
Incorporation into paints and coatings - 2-part
reactive coatings
Heritage IR, High BAF
1.91E02
4.62
4.35
3.91
4.07
4.12
Use in paints and coatings at job sites
Heritage IR, High BAF
1.11E02
2.69
2.53
2.27
2.37
2.40
Formulation of TCEP containing reactive resin
Heritage IR, High BAF
2.66E02
6.44
6.05
5.45
5.67
5.74
Laboratory chemicals
Heritage IR, High BAF
1.89
4.57E-02
4.30E-02
3.87E-02
4.03E-02
4.08E-02
Import and repackaging
Heritage IR, Low BAF
2.34
5.66E-02
5.33E-02
4.79E-02
4.99E-02
5.05E-02
Incorporation into paints and coatings -1-part
coatings
Heritage IR, Low BAF
1.04E01
2.52E-01
2.37E-01
2.13E-01
2.22E-01
2.25E-01
Incorporation into paints and coatings - 2-part
reactive coatings
Heritage IR, Low BAF
9.43
2.28E-01
2.15E-01
1.93E-01
2.01E-01
2.04E-01
Use in paints and coatings at job sites
Heritage IR, Low BAF
5.52
1.34E-01
1.26E-01
1.13E-01
1.18E-01
1.19E-01
Formulation of TCEP containing reactive resin
Heritage IR, Low BAF
1.32E01
3.19E-01
3.00E-01
2.70E-01
2.81E-01
2.85E-01
Laboratory chemicals
Heritage IR, Low BAF
9.41E-02
2.28E-03
2.14E-03
1.93E-03
2.00E-03
2.03E-03
1.5.5 Infant Risk Estimates
Table Apx 1-28. Infant Risks via Human Milk Exposure from Maternal Consumer Use Scenarios
COU Subcategory and Consumer Exposure Scenarios
Milk Intake
Rate Type
Intermediate
Chronic
Cancer
Fabric textile, leather products not covered elsewhere (carpet back coating)
Mean
26959
30129
2.85E-08
Fabric textile, leather products not covered elsewhere (textile for children's play structures)
Mean
1830
2045
4.19E-07
Building/construction materials not covered elsewhere (roofing insulation)
Mean
95
106
8.07E-06
Building/construction materials not covered elsewhere (acoustic ceiling)
Mean
1182
1321
6.49E-07
Foam seating and bedding product (foam automobile)
Mean
24083
26916
3.19E-08
Page 568 of 638
-------
COU Subcategory and Consumer Exposure Scenarios
Milk Intake
Rate Type
Intermediate
Chronic
Cancer
Foam seating and bedding product (foam living room)
Mean
10,771
12,037
7.12E-08
Foam seating and bedding product (mattress)
Mean
84,787
94,757
9.05E-09
Foam seating and bedding product (foam - other - toy block)
Mean
156,138
174,500
4.91E-09
Building/construction materials - Wood and engineered wood products (wood flooring)
Mean
109
122
7.04E-06
Building/construction materials - Wood and engineered wood products (wooden TV stand)
Mean
1,630
1,821
4.71E-07
Fabric textile, leather products not covered elsewhere (carpet back coating)
Upper
18,419
20,649
4.15E-08
Fabric textile, leather products not covered elsewhere (textile for children's play structures)
Upper
1,250
1,402
6.12E-07
Building/construction materials not covered elsewhere (roofing insulation)
Upper
65
73
1.18E-05
Building/construction materials not covered elsewhere (acoustic ceiling)
Upper
808
905
9.47E-07
Foam seating and bedding product (foam automobile)
Upper
16,455
18,447
4.65E-08
Foam seating and bedding product (foam living room)
Upper
7,359
8,250
1.04E-07
Foam seating and bedding product (mattress)
Upper
57,930
64,944
1.32E-08
Foam seating and bedding product (foam - other - toy block)
Upper
106,680
119,597
7.17E-09
Building/construction materials - Wood and engineered wood products (wood flooring)
Upper
74
83
1.03E-05
Building/construction materials - Wood and engineered wood products (wooden TV stand)
Upper
1,114
1,248
6.87E-07
Page 569 of 638
-------
TableApx 1-29. Infant Risks via Human Milk Exposure from Maternal Occupational Use Scenarios Based on Mean Milk Intake
Rate
OES
Route
Maternal Exposure
Duration
Intermediate
Chronic
Cancer
Import and repackaging
Dermal,
Inhalation
(High-end)
Chronic
1,054
1,178
7.28E-07
Incorporation into paints and coatings - 1-part coatings
Chronic
197
220
3.89E-06
Incorporation into paints and coatings - 2-part reactive coatings
Chronic
1,935
2163
3.97E-07
Processing - Formulation of TCEP into 2-part reactive resins
Chronic
956
1068
8.03E-07
Processing - Processing into 2-part resin article
Chronic
76
85
1.01E-05
Processing - Recycling Electronics
Chronic
1,206,274
1,348,128
6.36E-10
Commercial use - Paints and coatings - Spray (1-part, 250-day
application)
Chronic
114
127
6.74E-06
Commercial use - Paints and coatings - Spray (2-part reactive, 250-
day application)
Chronic
23
25
3.37E-05
Laboratory chemicals
Chronic
38
42
2.02E-05
Industrial/Commercial use - Installation of articles
Inhalation
(High-end)
Chronic
122,184,772
136,553,356
6.28E-12
Import and repackaging
Dermal,
Inhalation
(High-end)
Intermediate
89
99
8.64E-06
Incorporation into paints and coatings - 1-part coatings
Intermediate
28
32
2.71E-05
Incorporation into paints and coatings - 2-part reactive coatings
Intermediate
287
321
2.67E-06
Processing - Formulation of TCEP into 2-part reactive resins
Intermediate
101
113
7.59E-06
Processing - Processing into 2-part resin article
Intermediate
71
79
1.08E-05
Processing - Recycling electronics
Intermediate
1,126,657
1259,148
6.81E-10
Commercial use - Paints and coatings - Spray (1-part, 250-day
application)
Intermediate
106
119
7.21E-06
Commercial use - Paints and coatings - Spray (2-part reactive, 250-
day application)
Intermediate
21
24
3.61E-05
Laboratory chemicals
Intermediate
28
32
2.71E-05
Industrial/Commercial use - Installation of articles
Inhalation
(High-end)
Intermediate
114,120,319
127,540,532
6.72E-12
Page 570 of 638
-------
TableApx 1-30. Infant Risks via Human Milk Exposure from Maternal Occupational Use Scenarios Based on Upper Milk Intake
Rate
OES
Route
Maternal Exposure
Duration
Intermediate
Chronic
Cancer
Import and repackaging
Dermal,
Inhalation
(High-end)
Chronic
720
808
1.06E-06
Incorporation into paints and coatings - 1-part coatings
Chronic
135
151
5.68E-06
Incorporation into paints and coatings - 2-part reactive coatings
Chronic
1,322
1,482
5.78E-07
Processing - Formulation of TCEP into 2-part reactive resins
Chronic
653
732
1.17E-06
Processing - Processing into 2-part resin article
Chronic
52
58
1.48E-05
Processing - Recycling electronics
Chronic
824,172
923,965
9.28E-10
Commercial use - Paints and coatings - Spray (1-part, 250-day
application)
Chronic
78
87
9.83E-06
Commercial use - Paints and coatings - Spray (2-part reactive, 250-
day application)
Chronic
16
17
4.91E-05
Laboratory chemicals
Chronic
26
29
2.95E-05
Industrial/Commercial use - Installation of articles
Inhalation
(High-end)
Chronic
83,481,270
93,493,151
9.17E-12
Import and repackaging
Dermal,
Inhalation
(High-end)
Intermediate
61
59
1.45E-05
Incorporation into paints and coatings - 1-part coatings
Intermediate
19
19
4.56E-05
Incorporation into paints and coatings - 2-part reactive coatings
Intermediate
196
191
4.49E-06
Processing - Formulation of TCEP into 2-part reactive resins
Intermediate
69
67
1.28E-05
Processing - Processing into 2-part resin article
Intermediate
48
47
1.82E-05
Processing - Recycling electronics
Intermediate
769,775
748,688
1.15E-09
Commercial use - Paints and coatings - Spray (1-part, 250-day
application)
Intermediate
73
71
1.21E-05
Commercial use - Paints and coatings - Spray (2-part reactive, 250-
day application)
Intermediate
15
14
6.06E-05
Laboratory chemicals
Intermediate
19
19
4.55E-05
Industrial/Commercial use - Installation of articles
Inhalation
(High-end)
Intermediate
77,971,323
75,835,431
1.13E-11
Page 571 of 638
-------
TableApx 1-31. Infant Risks via Human Milk Exposure from Maternal General Population Oral Exposures Based on Mean Milk
Intake Rate
COUs/OESs
Route
Intermediate
Chronic
Cancer
Import and repackaging
Gen Pop Fish Ingestion, High BAF
259
290
2.96E-06
Incorporation into paints and coatings - 1-part coatings
Gen Pop Fish Ingestion, High BAF
59
65
1.31E-05
Incorporation into paints and coatings - 2-part reactive coatings
Gen Pop Fish Ingestion, High BAF
64
72
1.19E-05
Use in paints and coatings at job sites
Gen Pop Fish Ingestion, High BAF
110
123
6.97E-06
Formulation of TCEP containing reactive resin
Gen Pop Fish Ingestion, High BAF
46
52
1.66E-05
Laboratory chemicals
Gen Pop Fish Ingestion, High BAF
6,474
7,235
1.19E-07
Import and repackaging
Gen Pop Fish Ingestion, Low BAF
5,224
5,839
1.47E-07
Incorporation into paints and coatings - 1-part coatings
Gen Pop Fish Ingestion, Low BAF
1,179
1,318
6.51E-07
Incorporation into paints and coatings - 2-part reactive coatings
Gen Pop Fish Ingestion, Low BAF
1,300
1,453
5.90E-07
Use in paints and coatings at job sites
Gen Pop Fish Ingestion, Low BAF
2,219
2,480
3.46E-07
Formulation of TCEP containing reactive resin
Gen Pop Fish Ingestion, Low BAF
927
1,037
8.27E-07
Laboratory chemicals
Gen Pop Fish Ingestion, Low BAF
129,993
145,279
5.90E-09
Import and repackaging
Undiluted Drinking Water
5,224,384
5,838,756
1.47E-10
Incorporation into paints and coatings - 1-part coatings
Undiluted Drinking Water
1,179,218
1,317,891
6.51E-10
Incorporation into paints and coatings - 2-part reactive coatings
Undiluted Drinking Water
1,310,242
1,464,323
5.86E-10
Use in paints and coatings at job sites
Undiluted Drinking Water
2,224,940
2,486,586
3.45E-10
Formulation of TCEP containing reactive resin
Undiluted Drinking Water
932,715
1042,399
8.23E-10
Laboratory chemicals
Undiluted Drinking Water
131,024,263
146,432,358
5.86E-12
Page 572 of 638
-------
TableApx 1-32. Infant Risks via Human Milk Exposure from Maternal General Population Oral Exposures Based on Upper Milk
Intake Rate
COUs/OESs
Route
Intermediate
Chronic
Cancer
Import and repackaging
Gen Pop Fish Ingestion, High BAF
177
199
4.32E-06
Incorporation into paints and coatings - 1-part coatings
Gen Pop Fish Ingestion, High BAF
40
45
1.91E-05
Incorporation into paints and coatings - 2-part reactive coatings
Gen Pop Fish Ingestion, High BAF
44
49
1.74E-05
Use in paints and coatings at job sites
Gen Pop Fish Ingestion, High BAF
75
84
1.02E-05
Formulation of TCEP containing reactive resin
Gen Pop Fish Ingestion, High BAF
32
35
2.43E-05
Laboratory chemicals
Gen Pop Fish Ingestion, High BAF
4,423
4,959
1.73E-07
Import and repackaging
Gen Pop Fish Ingestion, Low BAF
3,569
4,002
2.14E-07
Incorporation into paints and coatings - 1-part coatings
Gen Pop Fish Ingestion, Low BAF
806
903
9.49E-07
Incorporation into paints and coatings - 2-part reactive coatings
Gen Pop Fish Ingestion, Low BAF
888
996
8.61E-07
Use in paints and coatings at job sites
Gen Pop Fish Ingestion, Low BAF
1,516
1,700
5.05E-07
Formulation of TCEP containing reactive resin
Gen Pop Fish Ingestion, Low BAF
634
710
1.21E-06
Laboratory chemicals
Gen Pop Fish Ingestion, Low BAF
88,816
99,570
8.61E-09
Import and repackaging
Undiluted Drinking Water
3,569,498
4,001,703
2.14E-10
Incorporation into paints and coatings - 1-part coatings
Undiluted Drinking Water
805,687
903,241
9.49E-10
Incorporation into paints and coatings - 2-part reactive coatings
Undiluted Drinking Water
895,207
1,003,602
8.54E-10
Use in paints and coatings at job sites
Undiluted Drinking Water
1,520,163
1,704,229
5.03E-10
Formulation of TCEP containing reactive resin
Undiluted Drinking Water
637,266
714,428
1.20E-09
Laboratory chemicals
Undiluted Drinking Water
89,520,729
100,360,171
8.54E-12
Page 573 of 638
-------
Table Apx 1-33. Infant Risks via Human Milk Exposure from Tribal Maternal Fish Exposures Based on Mean Milk Intake Rate
COUs/OESs
Route
Intermediate
Chronic
Acute Based on
Intermediate
Dose
Acute Based on
Chronic Dose
Cancer
Import and repackaging
Current IR, High BAF
27
30
92
103
2.89E-05
Incorporation into paints and coatings - 1-part coatings
Current IR, High BAF
6
7
21
23
1.28E-04
Incorporation into paints and coatings - 2-part reactive coatings
Current IR, High BAF
7
7
23
26
1.16E-04
Use in paints and coatings at job sites
Current IR, High BAF
11
13
39
44
6.79E-05
Formulation of TCEP containing reactive resin
Current IR, High BAF
5
5
16
18
1.62E-04
Laboratory chemicals
Current IR, High BAF
663
741
NA
NA
1.16E-06
Import and repackaging
Current IR, Low BAF
536
599
NA
NA
1.43E-06
Incorporation into paints and coatings - 1-part coatings
Current IR, Low BAF
121
136
NA
NA
6.32E-06
Incorporation into paints and coatings - 2-part reactive coatings
Current IR, Low BAF
133
149
NA
NA
5.76E-06
Use in paints and coatings at job sites
Current IR, Low BAF
228
254
NA
NA
3.37E-06
Formulation of TCEP containing reactive resin
Current IR, Low BAF
95
107
NA
NA
8.04E-06
Laboratory chemicals
Current IR, Low BAF
13,422
15,000
NA
NA
5.72E-08
Import and repackaging
Heritage IR, High BAF
3
4
12
14
2.20E-04
Incorporation into paints and coatings - 1-part coatings
Heritage IR, High BAF
1
1
3
3
9.76E-04
Incorporation into paints and coatings - 2-part reactive coatings
Heritage IR, High BAF
1
1
3
3
8.88E-04
Use in paints and coatings at job sites
Heritage IR, High BAF
1
2
5
6
5.16E-04
Formulation of TCEP containing reactive resin
Heritage IR, High BAF
1
1
2
2
1.24E-03
Laboratory chemicals
Heritage IR, High BAF
87
98
303
338
8.78E-06
Import and repackaging
Heritage IR, Low BAF
71
79
NA
NA
1.09E-05
Incorporation into paints and coatings - 1-part coatings
Heritage IR, Low BAF
16
18
55
61
4.83E-05
Incorporation into paints and coatings - 2-part reactive coatings
Heritage IR, Low BAF
18
20
61
68
4.38E-05
Use in paints and coatings at job sites
Heritage IR, Low BAF
30
33
NA
NA
2.57E-05
Formulation of TCEP containing reactive resin
Heritage IR, Low BAF
13
14
43
48
6.13E-05
Laboratory chemicals
Heritage IR, Low BAF
1,754
1,961
NA
NA
4.37E-07
Page 574 of 638
-------
Table Apx 1-34. Infant Risks via Human Milk Exposure from Tribal Maternal Fish Exposures Based on Upper Milk Intake Rate
COUs/OESs
Route
Intermediate
Chronic
Acute Based on
Intermediate
Dose
Acute Based on
Chronic Dose
Cancer
Import and repackaging
Current IR, High B AF
18
20
63
71
4.21E-05
Incorporation into paints and coatings - 1-part coatings
Current IR, High B AF
4
5
14
16
1.86E-04
Incorporation into paints and coatings - 2-part reactive coatings
Current IR, High B AF
5
5
16
18
1.69E-04
Use in paints and coatings at job sites
Current IR, High B AF
8
9
27
30
9.90E-05
Formulation of TCEP containing reactive resin
Current IR, High B AF
3
4
11
13
2.37E-04
Laboratory chemicals
Current IR, High B AF
453
508
NA
NA
1.69E-06
Import and repackaging
Current IR, Low BAF
366
411
NA
NA
2.09E-06
Incorporation into paints and coatings - 1-part coatings
Current IR, Low BAF
83
93
NA
NA
9.22E-06
Incorporation into paints and coatings - 2-part reactive coatings
Current IR, Low BAF
91
102
NA
NA
8.41E-06
Use in paints and coatings at job sites
Current IR, Low BAF
156
174
NA
NA
4.92E-06
Formulation of TCEP containing reactive resin
Current IR, Low BAF
65
73
NA
NA
1.17E-05
Laboratory chemicals
Current IR, Low BAF
9,170
10,281
NA
NA
8.34E-08
Import and repackaging
Heritage IR, High BAF
2
3
8
9
3.21E-04
Incorporation into paints and coatings - 1-part coatings
Heritage IR, High BAF
1
1
2
2
1.42E-03
Incorporation into paints and coatings - 2-part reactive coatings
Heritage IR, High BAF
1
1
2
2
1.30E-03
Use in paints and coatings at job sites
Heritage IR, High BAF
1
1
4
4
7.53E-04
Formulation of TCEP containing reactive resin
Heritage IR, High BAF
0
0
1
2
1.80E-03
Laboratory chemicals
Heritage IR, High BAF
60
67
NA
NA
1.28E-05
Import and repackaging
Heritage IR, Low BAF
48
54
NA
NA
1.59E-05
Incorporation into paints and coatings - 1-part coatings
Heritage IR, Low BAF
11
12
38
42
7.05E-05
Incorporation into paints and coatings - 2-part reactive coatings
Heritage IR, Low BAF
12
13
41
46
6.39E-05
Use in paints and coatings at job sites
Heritage IR, Low BAF
20
23
71
NA
3.74E-05
Formulation of TCEP containing reactive resin
Heritage IR, Low BAF
9
10
30
33
8.95E-05
Laboratory chemicals
Heritage IR, Low BAF
1,199
1,344
NA
NA
6.38E-07
Page 575 of 638
-------
1.5.6 Sensitivity Analysis
EPA conducted a sensitivity analysis for TCEP to evaluate the effect of chemical and biological
considerations on modeled TCEP concentrations in milk, as shown in Table Apx 1-35. Sensitivity was
measured using elasticity, which is defined as the ratio of percent change in each result to the
corresponding percent change in model input. A positive elasticity means that an increase in the model
parameter resulted in an increase in the model output, whereas a negative elasticity had an associated
decrease in the model output. FigureApx 1-7 shows the results of the sensitivity analysis.
Table Apx 1-35. Variables and Values Used in Sensitivity Analysis
Variable
Base/Default Values
Sensitivity Values
Half-Life
17.64
15.87, 19.40 (increased and decreased from base value
by 10%)
Kow17
60.26
66.28 and 54.23 (increased and decreased from base
value by 10%)
Lipid fraction in milk
0.038 + 0.000095*age
Multiplied the function by 1.1 and 0.9 to increase and
decrease from base value by 10%, respectively
Age at pregnancy
25
40 (increased to reflect an alternate scenario)
11 The analysis varied Kow rather than log Kow because the partition coefficient equations used are based on Kow.
Kow is not used elsewhere in the model equations.
-0.5 0 0.5 1 1.5
Half-life ^
Half-life 4/
KOW t
KOW 4/
Milk Lipid Fraction 'Is
Milk Lipid Fraction 4,
Figure Apx 1-7. Sensitivity Analysis of Model Inputs Measured as Elasticity
The elasticity for half-life is close to one. A ±10 percent change in half-life reflected a near equivalent
percent change in the infant milk dose. In contrast, a ±10 percent change to Kow resulted in a smaller
change in the infant milk dose. Half-life and Kow parameters are independent values in the model. The
half-life is used to estimate the liver compartment's elimination rate while Kow is used to estimate the
partition coefficients. For a slightly lipophilic compound like TCEP, an increase in Kow (and calculated
partition coefficient) leads to a relatively larger increase in the lipid:blood partition coefficient than for
Page 576 of 638
-------
other compartments such as mammary tissue. Thus, more TCEP will be stored in lipids and less in the
mammary tissue, causing a decrease in infant milk dose. If half-life increases, more TCEP is available in
the body and each compartment at a given time, including the mammary tissue, causing an increase in
infant milk dose. TCEP infant doses were insensitive to alterations of milk lipid fractions. TCEP
concentrations in milk were similarly insensitive (data not shown). This insensitivity may reflect the
relatively low Kow for TCEP.
Although the model treats Kow and half-life independently, these parameters are linked from a
toxicokinetic perspective. The Kow of the chemical likely influences both the partition coefficient (the
lipid compartments in particular) and the half-life. More lipophilic compounds tend to have larger
lipid:blood partition coefficient and longer half-lives than less lipophilic compounds. Thus, a 10 percent
change in Kow might also cause a percent change in the half life, and that correlation is not captured in
the model or sensitivity analysis.
Neither maternal age nor infant sex (results not shown) affected milk doses, indicating this model is not
sensitive to these parameters for TCEP. For infant sex, the only parameter differentiating male and
females in this model are growth curves, which are considered in the dose calculation.
Metabolic Rate
EPA conducted a similar sensitivity analysis for the metabolic rate of an infant compared to an adult.
The Verner model adjusts the intrinsic hepatic clearance rate such that the half-life of TCEP remains
constant throughout all lifestages irrespective of changes from birth through adulthood. This adjustment
does not consider that the metabolic potential of an infant is lower than an adult, and it is well known
that the levels of cytochrome P450 (CYP450) enzymes vary with age. Since the metabolic pathways for
TCEP is not fully characterized and it has a much shorter half-life than persistent organic pollutants (i.e.,
hours vs. years), parameterizing the Verner model to vary metabolic rates by age, body weight, or
another physiological factor could be speculative and not even result in noticeable differences.
Nonetheless, EPA still aimed to characterize CYP450 metabolism differences by varying TCEP's half-
life as a proxy.
The CYP450 metabolism of an infant was estimated at approximately 30 percent of an adult's based on
peer review comments and supported by (Ginsberg et al.. 2004). To be conservative, EPA assumed the
infant's metabolism remains at 30 percent throughout the first year of life, even though it increases
steadily during that time. TCEP's half-life was then increased by 3.3-fold (100/30) from 17.6 hours to
58.8 hours. EPA then estimated exposure and risks for worst-case scenarios: COUs with the highest
maternal doses and the upper milk intake rate. Both exposure and risk estimates for the nursing infant
still remain below that of the mother. The results are shown in TableApx 1-36 and TableApx 1-37:
Page 577 of 638
-------
Table Apx
-36. Average Infant Doses
Using a Longer TCEP Half-life
Population
Group
COU Subcategory
Route
Maternal
Exposure
Duration
Maternal
Dose
(mg/kg-
day)
Average
from Birth to
<1 Month
(mg/kg-day)
Average
from 1 to <3
Month
(mg/kg-day)
Average
from 3 to <6
Month
(mg/kg-day)
Average
from 6 to 12
Month
(mg/kg-day)
Overall Average
from Birth to 12
Month
(mg/kg-day)
Consumer
Building/construction materials
- wood and engineered wood
products (wood flooring)
Dermal,
Oral,
Inhalation
Chronic
1.80
1.45E-01
1.36E-01
1.22E-01
1.27E-01
1.29E-01
Occupational
Commercial Use - Paints &
Coatings - Spray (resin, 250-
day application 2-part coating)
Dermal,
Chronic
7.25
5.83E-01
5.48E-01
4.93E-01
5.12E-01
5.19E-01
Occupational
Commercial Use - Paints &
Coatings - Spray (resin, 250-
day application, 2-part coating)
Inhalation
Intermediate
7.76
6.24E-01
5.87E-01
5.28E-01
5.48E-01
5.56E-01
Tribal
Formulation of TCEP
containing reactive resin.
Current IR, High BAF
Chronic
3.49E01
1.62E01
1.52E01
1.37E01
1.42E01
1.44E01
Tribal
Formulation of TCEP
containing reactive resin.
Current IR, Low BAF
Oral
Chronic
1.73
1.39E-01
1.31E-01
1.18E-01
1.22E-01
1.24E-01
Tribal
Formulation of TCEP
containing reactive resin.
Heritage IR, High BAF
Chronic
2.66E02
2.14E01
2.01E01
1.81E01
1.88E01
1.91E01
Tribal
Formulation of TCEP
containing reactive resin.
Heritage IR, Low BAF
Chronic
1.32E01
1.06
9.98E-01
8.98E-01
9.32E-01
9.46E-01
Page 578 of 638
-------
Table Apx 1-37. Infant Risk Estimate Using a Longer TCEP Half-life
Population
Group
COU Subcategory and Consumer Exposure Scenarios
Intermediate
Chronic
Acute Based
on
Intermediate
Acute
Based on
Chronic
Cancer
Consumer
Building/construction materials - wood and engineered wood products
(wood flooring)
19
21
65
73
4.04E-05
Occupational
Commercial Use - Paints & Coatings - Spray (resin, 250-day application,
2-part coating)
5
5
16
18
1.63E-04
Occupational
Commercial Use - Paints & Coatings - Spray (resin, 250-day application,
2-part coating)
4
5
15
17
1.75E-04
Tribal
Formulation of TCEP containing reactive resin, Current IR, High BAF
0.2
0.2
0.6
0.7
4.52E-03
Tribal
Formulation of TCEP containing reactive resin, Current IR, Low BAF
20
22
68
76
3.89E-05
Tribal
Formulation of TCEP containing reactive resin, Heritage IR, High BAF
0.1
0.1
0.4
0.5
5.99E-03
Tribal
Formulation of TCEP containing reactive resin, Heritage IR, Low BAF
3
3
9
10
2.97E-04
Page 579 of 638
-------
1.6 Landfill Analysis Using DRAS
DRAS is an efficient tool developed by EPA Region 6 to provide a multipath risk assessment for the
evaluation of Resource Conservation and Recovery Act (RCRA) hazardous waste delisting. For the
TCEP Risk Evaluation, DRAS was specifically applied to model groundwater concentration estimates
from disposing TCEP to a hypothetical RCRA Subtitle D landfill at a range of loading rates and leachate
concentrations. A comprehensive description of the assumptions and calculations applied in DRAS can
be found in the Technical Support Document for the Hazardous Waste Delisting Risk Assessment
Software.
Because DRAS derives calculations based on a survey of drinking water wells located downgradient
from waste management units (U.S. EPA 19881 the model may provide the closest estimate to real
world scenarios available. Although there is some uncertainty inherent to applying the model as an
assessment tool under TSCA for risk evaluations, few other tools are available to effectively address this
pathway. This appendix will provide the input variables and calculations used to apply the model
determine potential groundwater concentrations. TableApx 1-38 and TableApx 1-39 provide the input
values used for each parameter in the model. Note that loading volumes were based on the range of
estimated production volumes (2,500 to 2,500,000 lb) and were calculated based on the density of TCEP
(1.39 g/cm3). For each loading volume, the range of leachate concentrations was applied.
Table Apx 1-38. Input Variables for Chemical of Concern
Input Variable for Chemical of Concern
Value
Chem Name
TCEP
CASRN
115-96-8
Maximum Contaminant Level
0
Oral Slope Cancer Factor
0.1t7
Inhalation Slope Cancer Factor (1/mg kg day)
0.018t7
Oral Reference Dose (mg/kg day)
0.03t?
Inhalation Reference Dose (mg/kg day)
0.03t?
Bioconcentration Factor (1/kg)
0
Soil Saturation Level
0
Toxicity Regulatory Rule regulatory level (mg/L)
Qc7
Henry's Law constant (atm -m3/mol)
2.95E-06
Diffusion Coefficient in Water (cm2/s)
5.07E-06
Diffusion Coefficient in Air (cm2/s)
0.04411
Water Solubility (mg/L)
7,820
Landfill Dilution Attenuation Factor
15.4
Surface Impoundment Dilution Attenuation Factor
3.18
Time to Skin Attenuation (hr/event)
0
Skin Permeability Constant (cm/hr )
0.0002211
Lag Time (hr)
0.28t?
Bunge Constant
4.1E-0517
Organic
Yes
Page 580 of 638
-------
Input Variable for Chemical of Concern
Value
Bioaccumulation Factor (L/kg)
6,016l1
Chronic Ecological Value (mg/L)
85l1
Carcinogen
No
Molecular Weight (g/mol)
285.49
Vapor Pressure (atm)
8.07E-5
Suspended sediment-surface water partitioning
coefficient (mg/L)
298.725
log Kow (log[mg/L])
1.78
Chemical Class
SVOC11
Analytical Method
8,260D11
Version Description
None11
Create Date
None11
Creator
None11
Cancer Risk Level
1.00E-0611
Hazard Quotient
ja
11 Input variables do not directly or indirectly affect groundwater concentrations
Table Apx 1-39. Waste Management Unit (WMU) Properties
Input Variable for WMU Properties
Value(s)
Waste Management Unit Type
Landfill
Loading Volume (m3)
8.17E-01
8.17
8.17E01
8.17E02
Cancer Risk Level
1.00E-06
Hazard Quotient
1.0
Detection Limit
0.5
Waste Management Active Life (years)
20
TCLP Concentration (mg/L)/Total
Concentration (mg/kg)
0.0001
0.001
0.01
0.1
1
10
100
1000
Page 581 of 638
-------
Once the model was executed for each loading rate and leachate concentration scenario, the groundwater
concentration was calculated using the leachate concentration and the 90th percentile weight-adjusted
dilution attenuation factor using
EquationApx 1-9.
r-..T Leachate Concentration
(j Wr — ,
Weight-Adjusted DAF
Where:
GWC = Groundwater concentration
Leachate concentration = Input variable for the waste management unit
Weight-AdjustedDAF= Weight- adjusted dilution attenuation factor
The results of these analyses are provided in Table 3-8.
Page 582 of 638
-------
Appendix J CONSUMER EXPOSURE DETAILS
J.1 Approach and Methodology
EPA evaluated TCEP exposure resulting from the use of consumer products and industrial processes.
The Agency utilized a modeling approach to evaluate exposure because chemical-specific personal
monitoring data attributable to the COUs was not identified for consumers during data gathering and
literature searches performed as part of systematic review using the evaluation strategies described in the
2021 Draft Systematic Review Protocol (U.S. EPA 2021a) and in the TCEP Systematic Review
Protocol (U.S. EPA 2024p).
There are a limited number of consumer articles that still contain TCEP, because many manufacturers
have reformulated them to remove TCEP. Consumer products containing TCEP are readily available via
the internet as finished articles (e.g., furniture and foam products). Use of these products can result in
exposures of the consumer user to TCEP during and after article use. Consumer exposure can occur via
inhalation, dermal, and oral routes.
Consumer products containing TCEP were identified through review and searches of a variety of
sources, including the National Institutes of Health (NIH) Household Products Database, various
government and trade association sources for products containing TCEP, company websites for safety
data sheets (SDSs), Kirk-Othmer Encyclopedia of Chemical Technology, and the internet. In general,
information on the consumer uses of TCEP was sparse and many manufacturers reported changes in
formulation and ceasing the use of TCEP in favor of other chemicals.
Identified consumer products (see Table l-l) were then categorized into six consumer use groups
considering (1) consumer use patterns, (2) information reported in SDSs, (3) product availability to the
public, and (4) potential risk to consumers.
Readers are referred to each model's user guide and associated user guide appendices for details on each
model, as well as information related to equations used within the models, default values, and the basis
for default values. Each model is peer reviewed. Default values within CEM are a combination of high-
end and mean or central tendency values derived from EPA's Exposure Factors Handbook (U.S. EPA.
2017d), literature, and other studies.
J.l.l Consumer Exposure Model (CEM)
CEM is a deterministic model that utilizes user provided input parameters and various assumptions (or
defaults) to generate exposure estimates. In addition to pre-defined scenarios, which align well with the
consumer uses identified in Table 1-1, CEM is peer reviewed, provides flexibility to the user allowing
modification of certain default parameters when chemical-specific information is available and does not
require chemical-specific emissions data (which may be required to run more complex indoor/consumer
models).
CEM predicts indoor air concentrations from consumer product use through a deterministic, mass-
balance calculation derived from emission calculation profiles within the model. There are six emission
calculation profiles within CEM (E1-E6) that are summarized in the CEM users guide and associated
appendices. If selected, CEM provides a time series air concentration profile for each run. These are
intermediate values produced prior to applying pre-defined activity patterns.
Page 583 of 638
-------
CEM uses a two-zone representation of the building of use when predicting indoor air concentrations.
Zone 1 represents the room where the consumer product is used. Zone 2 represents the remainder of the
building. Each zone is considered well-mixed. CEM allows further division of Zone 1 into a near field
and far field to accommodate situations where a higher concentration of product is expected very near
the product user when the product is used. Zone 1-near field represents the breathing zone of the user at
the location of the product use while Zone 1-far field represents the remainder of the Zone 1 room.
Inhalation exposure is estimated in CEM based on zones and pre-defined activity patterns. The
simulation run by CEM places the product user within Zone 1 for the duration of product use while the
bystander is placed in Zone 2 for the duration of product use. Following the duration of product use, the
user and bystander follow one of three pre-defined activity patterns established within CEM, based on
modeler selection. The selected activity pattern takes the user and bystander in and out of Zone 1 and
Zone 2 for the period of the simulation. The user and bystander inhale airborne concentrations within
those zones, which will vary over time, resulting in the overall estimated exposure to the user and
bystander.
CEM contains two methodologies for estimating dermal exposure to chemicals in products—the
permeability method (P-DER1) and the fraction absorbed method (A-DER1). Each of these
methodologies further has two model types, one designed for dermal exposure from use of a product (P-
DERla and A-DERla) and the other designed for dermal exposure from use of an article (P-DERlb and
A-DERlb). Each methodology has associated assumptions, uncertainties, and data input needs within
the CEM model. Both methodologies factor in the dermal surface area to body weight ratio and weight
fraction of chemical in a consumer product.
The permeability model is based on the ability of a chemical to penetrate the skin layer once contact
occurs. The permeability model assumes a constant supply of chemical, directly in contact with the skin,
throughout the exposure duration. The ability to use the permeability method can be beneficial when
chemical-specific skin permeability coefficients are available in the scientific literature. However, the
permeability model within CEM does not consider evaporative losses when it estimates dermal exposure
and therefore may be more representative of a dermal exposure resulting from a constant supply of
chemical to the skin due to a barrier or other factor that may restrict evaporation of the chemical of
interest from the skin such as a product soaked rag against the hand while using a product), or
immersion of a body part into a pool of product. Either of these examples has the potential to cause an
increased duration of dermal contact and permeation of the chemical into the skin resulting in dermal
exposure.
The fraction absorbed method is based on the absorbed dose of a chemical. This method essentially
measures two competing processes, evaporation of the chemical from the skin and penetration of the
chemical deeper into the skin. This methodology assumes the application of the chemical of concern
occurs once to an input thickness and then absorption occurs over an estimated absorption time. The
fraction absorbed method can be beneficial when chemical specific fractional absorption measurements
are available in the scientific literature. The consideration of evaporative losses by the fraction absorbed
method within CEM may make this model more representative of a dermal exposure resulting from
scenarios that allow for continuous evaporation and typically would not involve a constant supply of
product for dermal permeation. Examples of such scenarios include spraying a product onto a mirror and
a small amount of mist falling onto an unprotected hand. For TCEP, literature values for fraction
absorbed were used from Abdallah et al. (2016). rather than the faction absorbed estimation via CEM.
Page 584 of 638
-------
J. 1.2 Inputs
J.l.2.1 CEM and Sensitivity Analysis
Inputs for the each of the CEM 3.0 base and sensitivity runs are provide in Risk Evaluation for Tris(2-
chloroethyl) Phosphate (TCEP) - Supplemental Information File: Consumer Exposure Modeling Inputs
(U.S. EPA. 2024e). Where available, EPA relied on the Exposure Factors Handbook (U.S. EPA. 2017d)
and information identified in systematic review to inform input parameters. For article-specific
parameters (e.g., product density, thickness of article surface layer, surface area) that were unavailable
in the handbook or the peer-reviewed or gray literature, EPA used professional judgment to determine
whether the CEM default values were appropriate, or whether there should be an alternative value for
the parameter based on professional judgment. All the input parameters and their rationale are provided
in the Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental Information File:
Consumer Exposure Modeling Inputs (U.S. EPA. 2024e). Inputs for the sensitivity analysis are provided
in the "Sensitivity Analysis" tab of the Consumer Exposure Modeling Inputs (U.S. EPA. 2024e).
J. 1.3 Results
J.l.3.1 Raw Consumer Modeling Results
Modeling results are available in pdf and xlsx format in TCEP_Consumer_Modeling_Results.zip (U.S.
EPA. 2024b). Results from the consumer modeling have been visualized in bar charts, and risk tables in
the Navigating Supplemental Consumer Modeling Results Consumer Modeling Results were tabulated
in R and have been displayed in an "Rmarkdown file." The associated R script uses a workflow that
loads the input data from the consumer modeling results, cleans, filters, and wrangles the relevant data,
and displays the modeling results in the form of bar plots and risk tables.
Bar plots are interactive, and reviewers are able to pan and select certain data fields to help compare the
results from the various consumer COUs (see FigureApx J-l through FigureApx J-4). Hovering over
the data bars in the link above provides a tool tip that indicates the value of the bar.
Page 585 of 638
-------
Lifetime Average Daily Doses (LADDs)
TCEPCCjyS^ + p QQ
1.6
¦Eb
o
Q
0.8
5 °-6
0.4
0.2
0-1
¦
¦
J
Inhalation
Ingestion
Derma
Sfc, V %k_ %,* Nx, "V,
^ % \ H, ^
^v,,. " cou
%
* "X,
FigureApx J-l. Screenshot of Lifetime Average Daily Doses (LADDs) Bar
Chart Displaying Tool Tip for Acoustic Ceiling, Inhalation Estimate
The toolbar at the top also has various functionalities that can allow for more exploration of the data. For
example, simply hover and select the outlined double bars to compare data.
Lifetime Average Daily Doses (LADDs)
&b
TCEP COL":-,
t.s
1.6
0.6
23 Inhalaton
Qd
Inhalation
Ingestion
Dermal
V S,
cou
Figure Apx J-2. Screenshot of Lifetime Average Daily Doses (LADDs) Bar
Chart Displaying Function to Compare Data on Hover, for Insulation Estimates
Page 586 of 638
-------
Or to select and deselect data, the viewer can click the legend to remove data from the accompanying
bar plot.
Lifetime Average Daily Doses (LADDs)
TCEP COUs
0.4
0.3
M
&b 0 25
E
0.15
0.1
XnNVv
X cou
. Jj
A"
Irgestion
Derma!
FigureApx J-3. Screenshot of Lifetime Average Daily Doses (LADDs) Bar Chart Displaying Bar
Chart that Deselects Inhalation Estimate and Selects Ingestion and Dermal Estimates
Or the viewer can drag and select a certain section of the plot to view it in greater detail:
Lifetime Average Daily Doses (LADDs) Lifetime Average Daily Doses (LADDs)
TCEP COUs TCEP COUs
Figure Apx J-4. Screenshots of Lifetime Average Daily Doses (LADDs) Bar Chart Displaying a
Cropped Subsection of the Figure
J. 1.3.1 CEM 3.2 User Guide and Appendices
The CEM 3.0 User Guide and appendices provide the underlying equations and default parameters that are
used in CEM 3.2. The Risk Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental
Information File: Consumer Exposure Modeling Inputs (U.S. EPA. 2023) gives the inputs and
assumptions used for consumer modeling.
Page 587 of 638
-------
Appendix K HUMAN HEALTH HAZARD DETAILS
K.l Toxicokinetics and PBPK Models
K.1.1 Absorption
EPA did not identify in vivo human studies that evaluated absorption, distribution, metabolism, or
elimination (ADME) of TCEP by any route of exposure.
Oral
Following oral exposures to radiolabeled TCEP, in vivo ADME studies in rats and mice found that
TCEP is rapidly and extensively absorbed. More than 90 percent of 14C-labeled TCEP was absorbed
based on radioactivity found in urine, feces, volatiles, and CO2 after 2 hours post-dose (Burka et al..
1991; Herr et al.. 1991). In 5-week-old male Wistar rats, 14C-labeled TCEP concentrations were
measured in urine, feces, expired air, and body after oral exposure. Almost 100 percent of the 50
|iinol/kg dose was recovered in urine, feces, expired air, and body (Minegishi et al.. 1988). For input to
the risk evaluation, EPA will assume that absorption is 100 percent.
Inhalation
EPA did not identify any in vivo animal data for absorption of TCEP by the inhalation route of exposure.
For input to the risk evaluation, EPA will assume that absorption is 100 percent, equivalent to oral
exposure.
Dermal
EPA did not locate any in vivo studies of dermal absorption in humans or animals but identified an in
vitro study using excised human skin that evaluated the dermal absorption of TCEP (Abdallah et al..
2016).
Although no dermal in vivo toxicokinetic studies are available, EPA identified Abdallah et al. (2016).
which measured dermal absorption using excised human skin in multiple in vitro experiments conducted
according to OECD TG 428, Skin Absorption: In Vitro Method. The experiments used exposures of
either 24 or 6 hours; acetone or 20 percent Tween 80 in water as the vehicle; 500 or 1,000 ng/cm2
application to skin; and finite (depletable) or infinite dose. EPA gave each of the finite dose experiments
overall quality determinations of medium. For the experiment that claimed to investigate an infinite
dose, EPA assigned a low overall quality determination scenario, because conditions for infinite dosing
(use of neat or large body of material) were not met and the results did not reflect steady-state flux
throughout the experiment (e.g., applied dose was depletable).
EPA used the 500 ng/cm2 24-hour finite dose application in acetone (0.005 percent solution) to estimate
absorption for workers because this was the only experiment for which the authors reported absorption
at multiple time points. Because EPA assumes workers wash their hands after an 8-hour shift, EPA used
the value of 16.5 percent, which is the amount of TCEP absorbed at 8 hours. In accordance with OECD
Guidance Document 156 (OECD. 2022). EPA also added the quantity of material remaining in the skin
(6.8%) at the end of the experiment as potentially absorbable.47 Therefore, EPA assumes workers absorb
23.3 percent TCEP through skin and used this value to calculate risks for workers (see Section 5.1.1.3).
47 EPA used 6.8 percent (the total amount remaining in skin after washing) because the authors did not conduct tape
stripping.
Page 588 of 638
-------
For consumer exposures and exposure to soil scenarios that assume hand washing does not occur for 24
hours, EPA used the value at 24 hours (28.3%) plus the amount remaining in skin (6.8%) from the same
experiment used for workers (500 ng/cm2 24-hour finite dose application in acetone); total absorption
was 35.1 percent absorption and was used to calculate risks (see Sections 5.1.2.2.3 and 5.1.3.3.2).
The estimates identified above apply to finite exposure scenarios for which the TCEP dose is depleted
over time. For exposure scenarios such as swimming in which a maximum absorption rate is expected to
be maintained (i.e., the dose is not depletable during the exposure duration), EPA used the dermal
permeability coefficient (Kp) of 2.2/ 10 2 cm/h derived by Abdallah et al. (2016) from the experiment
that used the 24-hour 1,000 ng/cm2 TCEP skin application to calculate risks (see Section 5.1.3.3.1).
U.S. EPA (2024s) presents quality determinations for individual experiments conducted by Abdallah et
al. (2016). with EPA comments for each of the data quality metrics. Data extraction tables with details
on methods and results of the experiments are also presented in U.S. EPA (2024s).
K.1.2 Distribution
Oral
TCEP distributes widely throughout the body. At 2 hours following the oral exposure, there was TCEP-
derived 14C in all brain regions of male and female rats. Also, the increasing levels of TCEP-derived 14C
were observed with increasing TCEP doses. There were no significant differences in TCEP-derived 14C
levels in blood and brain (including cerebellum, brainstem, caudate, hypothalamus, cortex,
hippocampus, and midbrain) in male and female rats and 24 hours following a single dose. The
concentration of 14C-labeled TCEP in blood was significantly more increased with dose in males than
females after 2 hours (p < 0.05). However, there was no significant difference in the amount of TCEP
present in blood and all brain regions after 24 hours of exposure (Burka et al.. 1991; Herr et al.. 1991).
Oral administration studies in rats by NTP found that TCEP produced sex-specific seizures and lesions
in the hippocampal brain regions in some animals receiving the higher doses (NTP. 1991b). Results
reported by Herr et al. (1991) observed similar sex-specific clinical signs of toxicity in animals receiving
the higher doses. In an earlier study (Minegishi et al.. 1988). rats treated orally with TCEP had the
highest radioactivity in the kidney at 3, 6, and 12 hours, with similar or higher values in liver up to 168
hours (Minegishi et al.. 1988).
Inhalation
No in vivo animal data evaluating the distribution of TCEP following inhalation route exposures were
identified.
Dermal
EPA did not identify in vivo animal data that evaluated the distribution of TCEP following dermal route
exposures.
K.1.3 Metabolism
Oral
TCEP is predominantly metabolized in the liver in laboratory animals and urinary excretion is the
primary route of elimination for metabolites. In the liver, two pathways are involved in the metabolism
of TCEP (Burka et al.. 1991; Herr et al.. 1991). First pass biotransformation occurs via oxidative and
hydrolytic pathways. Some oxidative metabolites can undergo secondary biotransformation via the
glucuronidation and alcohol dehydrogenase pathways. Burka et al. (1991) conducted a study to detect
variations in metabolism of TCEP between male mice and male and female rats. The results showed that
TCEP underwent extensive metabolism in all three groups. TCEP was excreted primarily in the form of
Page 589 of 638
-------
metabolites in urine and feces of both species and were identified as bis(2-chloroethyl) hydrogen
phosphate (BCHP), the glucuronide of bis(2-chloroethyl) 2-hydroxyethyl phosphate (BCGP), and bis(2-
chloroethyl) carboxymethyl phosphate (BCCP) (Burka et al.. 1991). A synonym of BCHP is bis(2-
chlorethyl) phosphate (BCEP), which is used in various sections of this risk evaluation (e.g., 4.1.3.1
Measured Concentrations in Terrestrial Species and 5.1.3.5 Exposure Reconstruction Using Human
Biomonitoring Data and Reverse Dosimetry). In other toxicological studies in rats and mice, TCEP has
been shown to cause neurotoxicity at lower doses in females than in males (Yang et al.. 2018a; NTP.
1991b; Matthews et al.. 1990). Burka et al. (1991) examined whether there was any relationship between
acute neurotoxicity and metabolism. Male and female rats were pretreated with aldehyde dehydrogenase
inhibitors to alter the urinary metabolic profile. The relative amount of the hydrolytic metabolite
(BCHP) was increased compared to the oxidative metabolite (BCCP). Because aldehyde dehydrogenase
inhibitors interfere with the metabolic pathway leading to the oxidative metabolite (BCCP), increased
levels of the reactive metabolite may possibly account for increased neurotoxicity (Burka et al.. 1991).
Inhalation
No in vivo animal data for metabolism of TCEP by the inhalation route of exposure was identified.
Dermal
EPA did not identify in vivo animal data that evaluated metabolism of TCEP by the dermal route of
exposure.
K.1.4 Elimination
Oral
TCEP is primarily eliminated in the urine following oral exposure. Burka et al. (1991) and Herr et al.
(1991) reported that more than 75 percent of 14C4abeled TCEP was eliminated in 24 hours for both rats
and mice, with less than 10 percent excreted in feces (Burka et al.. 1991). There was little to no sex-
specific difference in the rate of elimination of TCEP for rats. However, male mice eliminated TCEP at
3 times the rate observed for rats during the first 8 hours (Burka et al.. 1991). Urinary excretion is the
primary route of elimination for metabolites (Burka et al.. 1991; Herr et al.. 1991). Minegishi et al.
(1988) also reported that almost 100 percent of 14C4abeled TCEP was eliminated, with 93.5 percent
excreted in urine.
Inhalation
No in vivo animal data for metabolism of TCEP by the inhalation route of exposure was identified.
Dermal
EPA did not identify in vivo animal data that evaluated elimination of TCEP by the dermal route of
exposure.
K.1.5 PBPK Modeling Approach
EPA did not identify any PBPK models specific to TCEP but is using the Verner Model (Verner et al..
2009; Verner et al.. 2008) to predict TCEP concentrations in milk used to assess infant exposure through
ingestion of human milk. The model is described in Appendix 1.5.1.
K.2 Detailed Mode of Action Information
EPA has determined that TCEP is likely to cause tumors in kidneys under exposure circumstances
relevant to human health. For blood cancer (mononuclear cell leukemia); thyroid cancer (follicular cell
adenoma or carcinoma); Harderian gland cancer (adenoma or carcinoma); and liver cancer
(hepatocellular adenomas or carcinomas), evidence of carcinogenicity is slight. EPA summarizes
Page 590 of 638
-------
biochemical, cellular, and mechanistic data that may be relevant to induction of kidney tumors—the
target organ with the strongest weight of scientific evidence conclusion.
Although EPA did not specifically investigate other possible mechanisms related to other tumor types
following TCEP exposure, conclusions for induction of kidney tumors may be relevant for induction of
other tumors.
K.2.1 Mutagenicity
EPA did not identify in vivo studies that evaluated any of the following relevant effects specifically in
kidneys, the target of tumors likely to be caused by TCEP: (1) oncogene or tumor suppressor gene
mutations, (2) other gene mutations and chromosomal aberrations, (3) DNA adducts, or (4) DNA
damage. However, one in vivo micronucleus assay in Chinese hamsters via intraperitoneal (i.p.)
administration did identify the presence of micronuclei in bone marrow (Sala et al.. 1982) and EPA
considered this to be equivocal/weakly positive.48 Also, the Agency did not identify any additional in
vivo studies that evaluated DNA damage, DNA adducts or other measures of DNA damage and/repair in
surrogate tissues.
Most bacterial reverse mutation assays using Salmonella typhimurium strains showed that TCEP was
negative for direct gene mutations (Follmann and Wober. 2006; NTP. 1991b; Haworth et al.. 1983;
Prival et al.. 1977; Simmon et al.. 1977). TCEP was also negative in a study of forward gene mutations
in Chinese hamster lung fibroblasts (Sala et al.. 1982).49
However, Nakamura et al. (1979) identified positive dose-response trends in two S. typhimurium strains:
in TA100, the response was less than 2-fold higher than the negative control at the highest non-toxic
dose, but in TA1535 (with metabolic activation), TCEP induced an increase of more 4- to 7-fold over
controls. It is not clear why the results of Nakamura et al. (1979) differed from other studies, but
Nakamura et al. (1979) used Kanechlor 500 to induce enzymes in the S9 fraction whereas other studies
used Aroclor 1254 or did not use a method to induce enzymes.
Two studies of TCEP induction of SCEs identified equivocal results in Chinese hamster ovary cells
(positive in one of two trials with S9, negative without S9) and positive results without a dose-response
in Chinese hamster lung fibroblasts (Galloway et al.. 1987; Sala et al.. 1982). suggesting some genetic
damage. These results are not definitive for direct mutagenic effects because there is a lack of
understanding of SCEs mechanism(s) of action (OECD. 2017).
TCEP was not considered to be an alkylating agent in an in vitro DNA binding assay (Lown et al..
1980).
Bukowski et al. (2019) conducted in vitro comet assays in peripheral mononuclear blood cells (PMBCs)
and identified DNA damage at the highest concentration tested (1 mM); however, there is uncertainty
regarding whether cytotoxicity occurred at this concentration. Another comet assay did not identify
DNA damage in Chinese hamster fibroblasts at TCEP concentrations up to 1 mM with or without
metabolic activation (Follmann and Wober. 2006).
48 Two additional micronucleus tests in mice (one via the oral route and one via i.p.) were negative (Beth-Hubner. 1999) but
the studies were not available for review by EPA.
49 Beth-Hubner (1999) reported negative results in a reverse gene mutation assay using Saccharomvces cerevisiae D4 and in
two mouse lymphoma assays (using the thymidine kinase locus).
Page 591 of 638
-------
Sala et al. (1982) identified a high level of cell transformation in Syrian hamster embryo (SHE) cells but
a lower level using C3H10T1/2 cells with metabolic activation. On page 24, OECD (2007) states that
"cell transformation has been related to structural alterations and changes in the expression of genes
involved in cell cycle control, proliferation and differentiation." The genomic changes may result from
direct or indirect genetic interactions or non-genotoxic mechanisms.
EPA did not identify in vitro studies of DNA adducts.
Although there is uncertainty regarding reasons for equivocal/weakly positive results, EPA concludes
that TCEP is not likely to induce tumors via a mutagenic MOA.
K.2.2 Other Modes of Action
Biochemical and mechanistic information that may suggest TCEP could act via MO As other than a
mutagenic MOA. Several in vivo and in vitro studies have evaluated tissue changes, gene transcription,
and protein activities among other activities that identified tumor precursors or possible key events in
mechanisms of tumor induction.
Taniai et al. (2012a) dosed male F344/NSIc rats daily via oral gavage with 0 or 350 mg/kg-bw/day
TCEP and examined effects on proximal tubular epithelial cells of the outer stripe of the outer medulla
(OSOM) of the kidney as well as the whole cortex. TCEP exposure resulted in scattered proximal
tubular regeneration, likely associated with cells in the quiescent GO-phase of the cell cycle. TCEP did
not induce karyomegaly (enlarged nuclei) in the tubular epithelia. TCEP also led to a significant increase
in Ki-67 immunoreactive cells vs. controls (p < 0.01); Ki-67 nuclear antigen is a marker of cell
proliferation expressed in cells in the G1 to M phase of the cell cycle. However, TCEP exposure did not
result in aberrant expression of cell cycle-related molecules except for topoisomerase Ila (Topo Ha),
which acts from the late S to G2 and M phase; TCEP significantly increased Topo Ila-immunoreactive
cells in the cortex and OSOM (p < 0.01), which may signify increased cell proliferation (Taniai et al..
2012a). It is also possible that DNA damage may have been a precipitating factor in the increase of
Topo Ila (Taniai et al.. 2012a).
Using the same protocol (i.e., male rats dosed via oral gavage at 0 or 350 mg/kg-day TCEP for 28 days),
Taniai et al. (2012b) observed that TCEP exposure increased cells immunoreactive for markers of cell
proliferation (Mcm3), apoptosis (Ubd), and deregulation of the G2/M phase of the cell cycle (TUNEL)
(p < 0.01). Carcinogens that increase cell proliferation may increase cell populations undergoing M
phase disruption that leads to chromosomal instability linked to cancer (Taniai et al.. 2012b).
In vitro studies show that TCEP exposure of primary rabbit renal proximal tubule cells (PTCs) resulted
in cytotoxicity, reduced DNA synthesis, altered expression of cell cycle regulatory proteins, and
inhibition of ion- and non-ion-transport functions. Increased expression of pro-apoptotic regulatory
proteins and decreased expression of proteins that inhibit apoptosis were also observed (Ren et al.. 2012;
Ren et al.. 2009. 2008).
Additional in vivo and in vitro studies identified several biochemical changes in tissues and cell of other
organs. Male ICR mice exposed to TCEP in the diet for 35 days exhibited increased markers of
oxidative stress (hepatic antioxidant enzyme activities and their gene expression) in livers (Chen et al..
2015a). Liver cells or cell lines cultured with TCEP exhibited reduced viability, cell cycle arrest, cellular
and mitochondrial oxidative stress, impaired mitochondrial function, and perturbation of cell signaling
pathways (Mennillo et al.. 2019; Zhang et al.. 2017b; Zhang et al.. 2017a; Zhang et al.. 2016c; Zhang et
al.. 2016b). TCEP exposure of human peripheral blood mononuclear cells resulted in cytotoxicity
(Mokra et al.. 2018) and decreased DNA methylation (Bukowski et al.. 2019).
Page 592 of 638
-------
In NTP (1991b). the authors reported no hyperplasia in rats at the 66-week interim sacrifice in the
narrative (data tables not included). Although focal hyperplasia was observed and can be expected to be
a precursor to tumors, the only related finding regarding kidney tumors at the 66-week sacrifice was a
single renal tubule adenoma seen in a female rat. Therefore, evidence of temporal progression from
hyperplasia to adenoma and then carcinoma is not available. At 2-years, hyperplasia was observed in
male rats but incidence was slightly lower (0, 2, and 24) than adenomas (1,5, and 24) compared with
hyperplasia at 0, 44, and 88 mg/kg-day. The lack of temporality and limited information on pre-cursor
lesions and their relationship with tumors leads to uncertainty regarding dose-response progression from
hyperplasia to adenomas and carcinomas in males. Female rats did have higher rates of hyperplasia (0,
3, 16) than adenomas (0, 2, 5), at 0, 44, and 88 mg/kg-day, respectively.
K.2.3 Mode of Action Conclusions
EPA concluded that a mutagenic MOA is not likely from exposure to TCEP. Several studies have
investigated biochemical and cellular changes in kidneys or renal cells that may be associated with steps
in other MO As for kidney cancer. However, EPA has not performed a formal analysis on postulated
MO As (e.g., as in Sonich-Mullin et al. (2001)).
There is sparse information on temporality and dose-response of potential pre-cursor events within the in
vivo studies and no clear NOAEL regarding tumor response to be able to model tumor incidence with a
non-linear/threshold dose response analysis.
U.S. EPA's PPRTV (U.S. EPA 2009) concluded that the overall weight of evidence for mutagenicity is
negative and that no mechanistic data identify specific potential key events in an MOA for kidney or
other tumors induced by TCEP exposure other than a general association with known proliferative and
preneoplastic lesions.
K.3 Dose-Response Derivation
EPA evaluated data for health outcomes with the strongest weight of scientific evidence and from
studies with sufficient sensitivity and adequate quantitative information to characterize the dose-
response relationships of TCEP (see Section 5.2.5.1).
K.3.1 Adjustments for All PODs (Non-cancer and Cancer)
For TCEP, all data considered for PODs are obtained from oral animal toxicity studies in rats or mice.
For consistency and easier comparison of sensitivity across health effects, EPA converted all doses to
daily doses before conducting benchmark dose (BMD) modeling. For example, if the toxicity study
dosed animals via gavage for five days per week at 22 mg/kg-day, EPA multiplied that value by 5/7 to
obtain an equivalent daily value of 15.7 mg/kg-day. Studies in which animals were dosed every day did
not require conversion. Any adjustments for different frequency of exposure (e.g., 5 days per week for
workers) are made in the exposure calculations specific to exposure scenarios.
Because toxicity values for TCEP are from oral animal studies, EPA must use an extrapolation method
to estimate equivalent human doses (HEDs) and CSFs. The preferred method would be to use chemical-
specific information for such an extrapolation. However, there are no TCEP-specific PBPK models and
EPA did not locate other TCEP information to conduct a chemical-specific quantitative extrapolation. In
the absence of such data, EPA relied on the guidance from U.S. EPA (2011b). which recommends
scaling allometrically across species using the three-quarter power of body weight (BW3/4) for oral data.
Allometric scaling accounts for differences in physiological and biochemical processes, mostly related
to kinetics.
Page 593 of 638
-------
For application of allometric scaling in risk evaluations, EPA uses dosimetric adjustment factors
(DAFs), which can be calculated using EquationApx K-l.
EquationApx K-l. Dosimetric Adjustment Factor (DAF)
U.S. EPA (2011b) presents DAFs for extrapolation to humans from several species. However, because
those DAFs used a human body weight of 70 kg, EPA has updated the DAFs using a human body
weight of 80 kg for the TCEP risk evaluation (U.S. EPA 2011a). The Agency used the body weights of
0.025 and 0.25 kg for mice and rats, respectively, as presented in U.S. EPA (2011b). The resulting DAFs
for mice and rats are 0.133 and 0.236, respectively.
For this risk evaluation, EPA assumes absorption for oral and inhalation routes is 100 percent and no
adjustment was made when extrapolating to the inhalation route. This is supported by oral toxicokinetics
data that shows greater than 90 percent absorption via the oral route (Burka et al.. 1991).
K.3.2 Non-cancer Dose-Response Modeling
EPA concluded that TCEP likely causes neurotoxicity, reproductive, developmental, and kidney effects
in humans under relevant exposure circumstances. For these outcomes (as well as suggestive evidence
integration conclusions), EPA conducted BMD modeling (U.S. EPA 2024c) and compared PODs
among these two categories of evidence integration conclusion categories to determine the sensitivity of
individual health affects (U.S. EPA 2024k). Although EPA conducted BMD modeling for the non-
cancer hazard outcomes with suggestive evidence integration conclusions, the focus of the evaluation
was on the likely endpoints. Section 5.2.5.1 describes how EPA chose the sensitive studies and
individual health effects within these health outcome categories for the non-cancer HED and HEC
derivations.
As noted above, EPA converted doses for each study to daily doses before conducting BMD modeling.
If data were not amenable to BMD modeling (e.g., there was only one treatment group) or data did not
fit BMD models, NOAELs or LOAELs were also converted to daily values, as needed.
Use of allometric scaling for oral animal toxicity data to account for differences among species allows
EPA to decrease the default intraspecies uncertainty factor (UFa) used to set the benchmark margin of
exposure (MOE); the default value of 10 can be decreased to 3, which accounts for any toxicodynamic
differences that are not covered by use of BW3 4. Using the appropriate DAF from Equation Apx K-l,
EPA adjusts the POD to obtain the daily HED as follows:
Equation Apx K-2. Daily Oral HED
Where:
DAF
BWa
BWh
Dosimetric adjustment factor (unitless)
Body weight of species used in toxicity study (kg)
Body weight of adult human (kg)
K.3.2.1 Calculating Daily Oral Human Equivalent Doses (HEDs)
HEDDaiiy — PODDauy X DAF
Where:
HED Daily
Human equivalent dose assuming daily doses (mg/kg-day)
Page 594 of 638
-------
POD Daily = Oral POD assuming daily doses (mg/kg-day)
DAF = Dosimetric adjustment factor (unitless)
K.3.2.2 Use of Oral HED as Dermal HED
U.S. EPA (2011b) recommends the BW3/4 approach only for oral PODs, and there is no established
guidance for dosimetric adjustments of dermal PODs. However, EPA only extrapolated between species
from oral animal toxicity values because the only acceptable data were from oral studies. EPA
extrapolated to the dermal HED from the oral HED after the oral species extrapolation and accounted for
differences in absorption in the dermal exposure estimate, not within the HEDs.
EPA used a value of 23.3 percent (hand washing after 8 hours) for workers as described in Section
5.1.1.3. EPA used a value of 35.1 percent (no handwashing for 24 hours) for dermal absorption in
calculations of consumer exposure and exposure to soil, which are described in Sections 5.1.2.2.3 and
5.1.3.3.2, respectively. For dermal exposure from swimming (a nondepletable source), EPA uses the
dermal permeability coefficient (Kp) of 2.2xlCT2 cm/h as described in Section 5.1.3.3.1. The same
uncertainty factors are used in the benchmark MOE for both oral and dermal scenarios.
K.3.2.3 Extrapolating to Inhalation Human Equivalent Concentrations (HECs)
For the inhalation route, EPA extrapolated the daily oral HEDs to inhalation human equivalent
concentrations (HECs) using a human body weight and breathing rate relevant to a continuous exposure
of an individual at rest, as follows:
EquationApx K-3. Extrapolating from Oral HED to Inhalation HEC
BWH
HECoaily, continuous ~ HEDj)aiiy X (— )
I tv ^ * Lj L/q
Where:
HECDaily,continuous = Inhalation HEC based on continuous daily exposure (mg/m3)
HEDoaiiy = Oral HED based on daily exposure (mg/kg-day)
BWh = Body weight of adult humans (kg) = 80
IRr = Inhalation rate for an individual at rest (m3/h) = 0.6125
EDc = Exposure duration for a continuous exposure (h/day) = 24
Based on information from U.S. EPA (2011a). EPA assumes an at rest breathing rate of 0.6125 m3/hr.
Adjustments for different breathing rates required for individual exposure scenarios are made in the
exposure calculations, as needed.
It is often necessary to convert between ppm and mg/m3 due to variation in concentration reporting in
studies and the default units for different OPPT models. Therefore, EPA presents all PODs in
equivalents of both units to avoid confusion and errors. Equation Apx K-4 presents the conversion of
the HEC from mg/m3 to ppm.
Equation Apx K-4. Converting Units for HECs (mg/m3 to ppm)
mg 24.45
X ppm = Y —5- x
m3 MW
Where:
Page 595 of 638
-------
24.45 =
MW =
Molar volume of a gas at standard temperature and pressure (L/mol), default
Molecular weight of the chemical
K.3.2.4 TCEP Non-cancer HED and HEC Calculations for Acute Exposures
Moser et al. (2015) identified neurotoxicity in pregnant female rats at 125 mg/kg-day via oral gavage in
a prenatal study. The POD is based on a NOAEL of 40 mg/kg-day (tremors within a few days of
dosing). EPA used EquationApx K-l to determine a DAF specific to rats (0.236), which was in turn
used in the following calculation of the daily HED using Equation Apx K-2:
mq mq
9.46 — = 40 — X 0.236
kg — day kg — day
EPA then calculated the continuous HEC for an individual at rest using Equation Apx K-3:
mq mq 80 kq
51.5 -f = 9.46- iL— x ( ^ )
m kg day 0.6125* 24 hr
hr
Equation Apx K-4 was used to convert the HEC from mg/m3 to ppm:
mg 24.45
4.41 ppm = 51.5 —- x —
HH m3 285
K.3.2.5 TCEP Non-cancer HED and HEC Calculations for Intermediate and Chronic
Exposures
Chen et al. (2015a) identified decreased numbers and degeneration of seminiferous tubules in male mice
in a 35-day study in which TCEP was administered in the diet. This endpoint is directly applicable to
intermediate exposure scenarios and because it is more sensitive than endpoints from the chronic
studies, EPA also uses it for chronic exposure scenarios. The POD is based on a BMDLs of 21.0 mg/kg-
day. EPA used Equation Apx K-l to determine a DAF specific to rats, which was in turn used in the
following calculation of the daily HED using Equation Apx K-2:
mq mq
2.79—— = 21.0—— X 0.133
kg kg
EPA then calculated the continuous HEC for an individual at rest using Equation Apx K-3:
_ 80 kg
15.2 mg/m = 2.79 mg /kg x ( ^ )
0.6125?-* 24 hr
hr
Equation Apx K-4 was used to convert the HEC from mg/m3 to ppm:
mg 24.45
1.30 ppm = 15.2 —7 x
m3 285
K.3.3 Cancer Dose-Response Modeling
EPA concludes that TCEP is likely to be carcinogenic to humans based on considerations outlined in
U.S. EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA. 2005b). EPA modeled the dose
Page 596 of 638
-------
response for the target organ with the most robust data—kidney tumors. For tumors in several other
target organs, see the evidence integration tables in Appendix L.
K.3.3.1 Calculating Daily Oral CSFs
Like non-cancer data, all cancer data are obtained from oral animal toxicity studies (NTP. 1991b).
Because an MOA has not been established for TCEP, EPA assumed linear low dose extrapolation (U.S.
EPA. 2005b). EPA conducted BMD modeling of kidney tumors for both male and female rats to obtain
the CSF for TCEP (U.S. EPA 2024c). EPA adjusted the CSF using the DAF (see EquationApx K-l) to
account for allometric scaling between species. Equation Apx K-5 shows the calculation to obtain the
DAF-adjusted CSF:
Equation Apx K-5.
Where:
CSF Human, Daily
CSFAnimal, Daily
DAF
CSF,
Human,Daily
= CSF,
Animal
.Daily /DAF
Human equivalent daily oral cancer slope factor (mg/kg-day ')
Animal daily oral cancer slope factor (mg/kg-day ')
Dosimetric adjustment factor (unitless)
Because EPA has not concluded that TCEP acts via a mutagenic MOA, an age-dependent adjustment
factor (ADAF) (U.S. EPA 2005c) was not applied. The Agnecy did not use CSFs for combined tumors
(across multiple target organs) for the risk evaluation but focused on the tumors with the most robust
evidence from the animal data.
K.3.3.2 Use of Oral CSF as Dermal CSF
The BW3/4 approach is only recommended for oral toxicity data extrapolation, and there is no established
guidance for dosimetric adjustments of dermal PODs. In the absence of available guidance, and when
the dermal CSFs are extrapolated from oral CSFs that incorporated BW3/4 scaling, EPA uses the oral
CSF for the dermal route of exposure because it has already been converted to a human dose. The
Agency accounts for dermal absorption in the dermal exposure estimate, which can then be directly
compared to this FLED. Sections 5.1.2.2.3 and 5.1.3.3.2 describe how EPA uses dermal absorption in
calculations of consumer exposure and exposure to soil, respectively; Section 5.1.1.3 describes dermal
exposure for workers; and Section 5.1.3.3.1 describes dermal exposure from swimming (an infinite,
nondepletable source).
K.3.3.3 Extrapolating to Inhalation Unit Risks (IURs)
For the inhalation route, EPA extrapolated the daily oral HEDs to inhalation HECs using a human body
weight and breathing rate relevant to a continuous exposure of an individual at rest. For this risk
evaluation, EPA assumes absorption for oral and inhalation routes is equivalent and no adjustment was
made when extrapolating from the oral to the inhalation route. The equation to convert to the inhalation
route is as follows:
Equation Apx K-6. Extrapolating from the Oral CSF to an Inhalation IUR
urn rcr? JRr*EDc^
IURnuman,continuous ~ ^^^Human,daily ^ ^
Where:
IURHuman, continuous ~ Human equivalent continuous daily inhalation unit risk ((mg/m3) ')
CSF Human, daily = Human equivalent daily oral cancer slope factor (mg/kg-day ')
Page 597 of 638
-------
IRr = Inhalation rate for an individual at rest (m3/hr) = 0.6125
EDc = Exposure duration for a continuous exposure (hr/day) = 24
BWh = Body weight of adult humans (kg) = 80
Based on information presented in U.S. EPA (2011a). EPA assumes an at rest breathing rate of 0.6125
m3/hr.
EPA may need to convert between mg/m3 and ppm due to variation in concentration reporting in studies
and the default units for different OPPT models. Therefore, all PODs are presented in equivalents of
both units to avoid confusion and errors. EquationApx K-7 identifies how to convert the IUR from
(mg/m3)-1 to (ppm)-1.
Equation Apx K-7. Converting Units for IURs (mg/m3 to ppm)
mg MW
X per ppm = Y per —7 x
ms 24.45
Where:
24.45 = Molar volume of a gas at standard temperature and pressure (L/mol), default
MW = Molecular weight of the chemical
K.3.3.4 CSF and IUR Calculations for Lifetime Exposures
The most sensitive CSF was estimated as a risk of 0.0058 per mg/kg-day using BMD modeling software
to model the dose-response for renal tubule adenomas and carcinomas in male rats from the NTP
(1991b) 2-year cancer bioassay. EPA then used this CSF and the rat-specific DAF (0.24) (Equation Apx
K-l) to obtain a human relevant CSF using Equation Apx K-5. The calculations specific to TCEP are as
follows:
mq mq
0.0245 per—^ = 0.0058 per—^ /0.236
kg kg
Using Equation Apx K-6, EPA converted the oral CSF to an IUR:
mg 0.6125 m3/hr * 24 hr
0.00451 per = 0.0245 per mg/kg x ( )
ms 80 kg
EPA used Equation Apx K-7 to convert the IUR from units of mg/m3 to ppm:
mg 285
0.0526 per ppm = 0.00451 per —7 x
ms 24.45
Page 598 of 638
-------
Appendix L EVIDENCE INTEGRATION FOR HUMAN HEALTH
OUTCOMES
This appendix presents evidence integration tables for the major health outcomes associated with TCEP
(see TableApx L-l through TableApx L-6). It also presents a section with short evidence integration
summaries for health outcomes with limited data (Section L.2).
Page 599 of 638
-------
L.l Evidence Integration Tables for Major Human Health Hazard Outcomes
Table Apx L-l. Evidence Integration for Neurotoxicity
Database Summary
Factors that Increase
Strength
Factors that
Decrease Strength
Summary of Key
Findings and Within-
stream Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
Evidence integration summary judgment on neurotoxicity
Evidence in studies of exposed humans considered for deriving toxicity values
Overall judgment for
neurological/
behavioral effects based
on integration of
information across
evidence streams:
Evidence indicates that
TCEP likely causes
neurological/
behavioral effects in
humans under relevant
exposure circumstances.
• Associations between BCEP urine levels (metabolite
of TCEP) in ore a nancy and birth outcomes:
Percv et al. (2021) examined whether orenatal
exposure of suspected neurotoxicants was associated
with any child cognition measures. Maternal urinary
BCEP was associated with a modest increase in child
full scale IQ.
Percv et al. (2022) examined impairment in cognitive
abilities a longitudinal cohort study. Urinary BCEP
concentrations at ages 1-5 years and cognitive abilities
at 8 years were assessed for in. The BCEP
concentrations association with IQ was small and
positive.
Hernandez-Castro et al. (2023a) evaluated associations
of prenatal exposures to organophosphate esters and
child neurobehavior. The authors did not find an
association between urinary BCEP levels and
neurobehavioral outcomes.
Foster et al. (2024) examined whether oreenant women
exposed to organophosphate esters that also included
tris(2-chloroethyl) phosphate (TCEP) in home dust was
associated with higher depression and stress levels
across prenatal and postpartum time periods. The
authors observed small increases in maternal perceived
stress levels, but not depression, after adjusting for
covariates.
Consistency:
Prenatal exposure to TCEP
is associated with
impairment in cognitive
abilities for multiple
measures of low SES.
Biolosical olausibi 1 it\ and
relevance to humans:
Neurodevelopmental
effects measured in
humans
Consistency:
Increase and decrease
of IQ
The epidemiological
studies evaluated
participants' single
spot urine samples
Key findings:
Available
epidemiological studies
resulted in some changes
in cognitive functions in
children associated with
maternal urine
concentrations of BCEP
as well as TCEP in home
dust.
Overall judgment for
neurological effects based
on human evidence:
• Slight
Evidence from in vivo mammalian animal studies considered for deriving toxicity values
Page 600 of 638
-------
Database Summary
Factors that Increase
Strength
Factors that
Decrease Strength
Summary of Key
Findings and Within-
stream Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
NTP studies (Matthews et al.. 1993; NTP. 1991b:
Matthews et al.. 1990). Rats and mice exposed bv
gavage; evaluated brain/hippocampal lesions, clinical
signs of toxicity, serum cholinesterase activity. Overall
quality determination: High
Brain/hiDDOcaniDal lesions (historatholoev) (16 weeks,
and two years [rats only])
• Female rats: brain weight decrease observed at the
highest dose.
• Male rats: necrosis of the neurons of the
hippocampus,
• Female rats: necrosis of the neurons of the
hippocampus. Neuronal necrosis was also observed
in the thalamus.
• Female rats: in over 40% of female rats receiving
the highest dose showed focal gliosis, hemorrhage,
mineralization, and pigmentation, and hemosiderin
in the brain stem and cerebellum after 2 years.
Clinical sisns of toxicity (16 davs. andl6 weeks)
• Female rats: occasionally appeared hyperactive and
exhibited resistance to handling. Seizures were
observed during week 12 of dosing.
• Male rats: no clinical signs of toxicity were
observed in male rats.
• Male and female mice exhibited convulsive
movements and reduced ability to keep balance
during the first three days of dosing at the two
highest doses.
Serum cholinesterase activity
• Female rats: serum cholinesterase activity was
decreased at the highest doses after 14 days.
• Female rats: serum cholinesterase activity in female
rats receiving the higher were 75% and 59%,
respectively, of the control animals. The 88
Effect sizc/orecision:
• Histopathology, serum
cholinesterase activity,
behavioral changes in
female rats were
significantly increased
over controls.
Dosc-rcsDonsc eradient:
• Decrease in serum
cholinesterase activity
appears to increase with
dose in female rats.
Incidences of brain
histopathology findings
increased with dose in
male and female rats.
Consistency:
• Brain weight,
brain/hippocampal
lesions, clinical signs of
toxicity, serum
cholinesterase activity,
and behavioral findings
were observed in
female rats across
different studies.
Coherence across
endDoints:
• Signs of neurotoxicity
and neurobehavioral
effects corresponded
to histopathology
changes in female
rats.
Consistency:
• Effects seen
primarily in female
rats in some
studies.
• Neurodevelopment
al limited effects
seen only in male
offspring at the
highest dose but
only in the
presence of severe
maternal toxicity.
Key findings'.
Results across available
animal toxicological
studies showed
neurotoxicity in female
rats in a dose-response
manner and limited
effects in male rat
offspring showed effects
at the highest dose in a
prenatal study.
Overall judgment for
neurotoxicity based on
animal evidence:
• Robust
Page 601 of 638
-------
Database Summary
Factors that Increase
Strength
Factors that
Decrease Strength
Summary of Key
Findings and Within-
stream Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
mg/kg/day animals were decreased 9.3% compared
to control animals.
• There were no treatment-related effects on serum
cholinesterase activity in both male and female
mice
Tilson et al. (1990). 1-dav savase studv in rats:
evaluated hippocampal lesions and behavioral
findings. Overall quality determination: High
• Treatment produced consistent damage to CA1
pyramidal cells with lesser damage to CA4, CA3,
and CA2 pyramidal cells. Significant damage was
also seen in dentate granule cells.
• Treated rats were mildly impaired in the acquisition
of the water maze task that had a reference memory
component. However, in the repeated acquisition
task, the rats were clearly deficient.
Yans et al. (2018a). 60-dav savase studv in rats:
evaluated clinical signs of toxicity hippocampal
lesions, and behavioral findings. Overall quality
determination: High
Clinical sisns of toxicitv
• Occasional periods of hyperactivity and periodic
convulsions in female rats. There were not
treatment-related effects observed in male rats
Behavioral findinss
• Remarkably higher escape latencies to find the
hidden platform than the vehicle controls (p <
0.01). Significantly shorter cumulative distances
from the original platform than the controls.
Significantly fewer cross-times were noted in the
highest dose for female rats. Male rats were not
tested.
Page 602 of 638
-------
Database Summary
Factors that Increase
Strength
Factors that
Decrease Strength
Summary of Key
Findings and Within-
stream Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
Hazleton Laboratories (1983). a sinele dose durine GD
7-14. Overall quality determination: High
• There was a low incidence of maternal animals
with clinical signs of OP toxicity (up to 2/50
animals on GD 7-14).
Developmental Neurotoxicity:
Moser et al. (2015) Overall aualitv determination:
High
Assessment of neurobehavioral and related hormonal
responses after dosing pregnant Long-Evans rats from
GD 10 through PND 22 via oral gavage up to 90
mg/kg-day. No TCEP-related adverse effects in T3,
T4, brain or serum ACliE in dams or offspring. In
addition, no effects on brain weight in offspring at
PND 6 and sporadic behavioral changes do not suggest
biologically relevant adverse outcomes or
developmental toxicity.
Kawashima et al. (1983) Overall aualitv determination:
Medium
Pregnant Wistar rats gavaged with 0, 50, 100, or 200
mg/kg-day from GD 7 through 15. Twenty-tliree
percent of 30 dams at 200 mg/kg-day died between GD
10 and 14. At 200 mg/kg-day, male offspring exhibited
decreased numbers of rearings (9.8 vs. 19.3 in controls;
p < 0.01) and took longer during the learning ability
test (water maze performance) in the last of four trials
(p < 0.05). However, this dose group was associated
with significant maternal toxicity. No effects were
observed in female offspring in these tests or in either
sex in other functional/neurobehavioral tests.
Evidence in mechanistic studies and supplemental information
In vivo:
Yans et al. (2018a). Compared to those in the control,
the major metabolites that had increased in the aqueous
• None
• None
Overall judgment for
neurotoxicity based on
mechanistic evidence:
Page 603 of 638
-------
Database Summary
Factors that Increase
Strength
Factors that
Decrease Strength
Summary of Key
Findings and Within-
stream Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
phase of TCEP-treated groups were N-acetyl aspirate
(NAA), glutamine (GLU), glutamic acid, glucose,
taurine, choline, creatine, and myo-inositol levels,
whereas those that had decreased were lactate, g-amino
butyric acid (GABA), glycine, and two unknown
compounds. In the lipid phase, the major metabolites
that were different between the control and TCEP-
treated groups were cholesterol ester and glycerol,
which were increased, whereas free cholesterol, total
cholesterol, lipid (CH2CH2CO), fatty acid,
polyunsaturated fatty acid, and phosphatidylcholine
levels were decreased.
• Indeterminate
Page 604 of 638
-------
Table Apx L-2. Evidence Integration for Reproductive Effect
S
Database Summary
Factors that Increase Strength
Factors that Decrease Strength
Summary of Key
Findings and
Within-stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
Evidence integration summary judgment on reproductive effects
Evidence in studies of exposed humans considered for deriving toxicity values (none)
Overall judgment for
reproductive effects
based on integration of
information across
evidence streams:
Evidence indicates that
TCEP exposure likely
causes reproductive
effects in humans under
relevant exposure
circumstances.
Evidence from apical endpoints in in vivo mammalian animal studies considered for deriving toxicity values
• Short-term, subchronic, and
chronic gavage studies in male and
female rats and mice and a
subchronic dietary study in male
mice examined testes weight and/or
histology of the reproductive
orsans NTP (1991b) and Chen et
al. (2015a). Overall aualitv
determination: High
• The Reproductive Assessment by
Continuous Breeding (RACB)
Protocol5" was used to evaluate
fertility, litters/pair, live pups/litter,
proportion of pups born alive, sex
of live pups, pup weights at birth,
sperm morphology, vaginal
cytology, and/or reproductive organ
weights and histology in mice
treated via savase (NTP. 1991a).
Overall quality determination:
High.
• Biolosical eradient/dose-
rcsdorise: The magnitude and
severity of histological changes
in the testes (changes in the
number and appearance of
seminiferous tubules)
increased with increasing dose
in the subchronic dietary study
in ICR mice.
• Fertility index, number of
litters/pair decreased in a dose-
related manner during the
continuous FO breeding phase
of the RACB.
Consistencv:
• Decreased testes weight was
observed in gavage and dietary
subchronic studies in mice.
• Decreased fertility index was
observed during continuous FO
breeding and crossover mating
phases of the RACB.
• Sperm effects (decreases on
sperm concentration and
percent motile sperm,
increased sperm abnormalities)
identified during crossover
mating correlated with
Consistencv:
• Changes in testes histology were
observed in a subchronic dietary
study in ICR mice, but no
histological changes to
reproductive organs were
observed in short-term,
subchronic, or chronic gavage
studies in F344 rats and CD-I
andB6C3Fl mice.
Oualitv of the database:
• Testes weights were assessed in
subchronic, but not chronic, NTP
studies in rats and mice.
Key findings'.
Available animal
toxicological studies
showed decreased
testes weight,
histological changes
in the testes of ICR
mice, sperm effects,
and/or reduced
fertility and
fecundity.
Overall judgment for
reproductive effects
based on animal
evidence:
• Moderate
511 The RACB protocol consists of 4 phases: (1) dose range-finding, (2) continuous (FO) breeding, (3) crossover mating; and (4) assessment of fertility in F1 offspring.
Page 605 of 638
-------
Database Summary
Factors that Increase Strength
Factors that Decrease Strength
Summary of Key
Findings and
Within-stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
decreased fertility index when
treated males were bred with
untreated females.
• Mechanistic changes from in
vivo and in vitro studies
(decreased testicular
testosterone, altered gene
expression related to
steroidogenesis, and decreased
testosterone secretion) are
consistent with observed
effects on testes and sperm.
Oualitv of the database:
• Effects were observed in high-
quality studies.
Evidence in mechanistic studies and supplemental information
• A subchronic dietary study in male
mice evaluated testicular
testosterone and gene expression
related to testosterone synthesis
(Chen et al.. 2015a).
• An in vitro study using TM3
Leydig cells evaluated testosterone
secretion and gene expression
related to steroidogenesis and
oxidative stress (Chen et al..
2015b).
• Three in vitro studies evaluated
estrogenic, anti-estrogenic,
androgenic, and/or anti-androgenic
activity using a yeast reporter assay
or human (endometrial, prostate
and breast) cancer cell lines
(Krivoshiev et al.. 2016; Reers et
al.. 2016; Follmann and Wober.
Biolosical sradient/dose-
rc sdo rise:
• In vivo data showed decreased
testicular testosterone and
altered gene expression related
to testosterone synthesis at the
dose in which decreased testes
weight and testicular damage
were observed.
• An in vitro study showed
decreased testosterone
secretion and/or changes in
gene expression related to
steroidogenesis and oxidative
stress at both tested
concentrations.
Consistency:
• Altered gene expression
related to steroidogenesis
Consistency:
• There was inconsistency across
studies with respect to estrogen
receptor and androgen receptor
agonist and/or antagonist activity
in human (endometrial, prostate,
and breast) cancer cell lines.
Oualitv of the database:
• Few potential mechanisms were
investigated in available studies.
Biolosical olausibilitv/relevance to
humans:
• Oxidative stress is a nonspecific
mechanism.
Key findings: Limited
available mechanistic
data indicate that
TCEP may induce
oxidative stress and
endocrine disruption
via altered expression
of genes involved in
steroidogenesis.
Overall judgment for
reproductive effects
based on mechanistic
evidence:
• Slight
Page 606 of 638
-------
Database Summary
Factors that Increase Strength
Factors that Decrease Strength
Summary of Key
Findings and
Within-stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
2006).
correlated with decreased
testosterone in vivo and in
vitro.
Biolosical olausibilitv/relevance
to humans:
• Endocrine disruption, via
altered expression of genes
involved in testosterone
synthesis, is a plausible
mechanism for infertility,
sperm effects, and testicular
damage that is relevant to
humans.
GD = gestation day
Page 607 of 638
-------
Table Apx L-3. Evidence Integration for Developmental Effects
Database Summary
Factors that Increase Strength
Factors that Decrease Strength
Summary of Key
Findings and within-
Stream Strength of
the Evidence
Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
Evidence integration summary judgment on developmental effects
Evidence in studies of exposed humans considered for deriving toxicity values (none)
Overall judgment for
developmental effects
based on integration of
information across
evidence streams:
Evidence suggests but is
not sufficient to
conclude that TCEP
causes developmental
effects in humans under
relevant exposure
circumstances.
• Associations between BCEP urine
levels (metabolite of TCEP) in
oreenancv and birth outcomes:
-U.S. cohort of 6,646 mother/ infant
pairs that investigated gestational
age (GA), birthweight (BW) (Oh
et al.. 2024) Overall aualitv
determinations: Medium
-Los Angeles, CA cohort of 421
mother/infant pairs - evaluation
of GA. BW (Hernandez-Castro et
al.. 2023b) Overall aualitv
determinations: Medium
- Cincinnati, Ohio cohort of 340
mom/infant pairs and GA, pre-
term birth, BW, length ponderal
index, head circumference (Yans
et al.. 2022) Overall aualitv
determinations: Medium
- Rhode Island cohort 56
mom/infant pairs and skinfold
thickness, BW, length, head and
abdominal circumferences,
feeding behavior at birth and 6
wks (Crawford et al.. 2020)
Overall quality determinations:
High
Quality of the database:
High/medium
Consistencv: Males had lower
incidence of being small for their
gestational age and had increased
scapular skinfold thickness,
which could both suggest some
increase in fat. Two studies
showed statistically significant
effects on gestational age and/or
birthweight and length (increased
GA for both sexes, decreased for
females).
Biolosical sradient/dose-
rcsDonsc:
Individual measures showed
dose-response gradients;
Yans et al. (2022) identified
differences in multiple
measures for each log 10 unit
increase in BCEP
Crawford et al. (2020) showed
increases in some skinfold
thickness measures.
Biolosical olausibilitv and
relevance to humans: The tvDcs
of outcomes are seen in humans
across a variety of populations.
Consistencv: One of the larser
cohorts (Hernandez-Castro et al..
2023b) did not show effects on
gestational age or birth weight.
Others showed inconsistent effects
between males, females, and when
both sexes were evaluated together.
Crawford et al. (2020) also showed
only changes in some skinfold
thickness measures but not infant
length or weight, whereas females
showed changes in weight and
lensth in another studv (Yans et al..
2022).
Differences in effects among sexes
were observed.
Biolosical sradicnt/dosc-rcsDonsc:
The dose-response was not always
maintained across high, low, and
non-detect categories (e.g., SGAfor
males, which showed a slightly
greater change between the low
dose group and controls vs. high
dose and controls for males) and the
females showed only a statistically
significant change for low dose vs.
control for LGA - large for
sestational ase (Oh et al.. 2024).
Key findings'.
Available
epidemiological
studies resulted in
some changes in
gestational age and
growth.
Overall judgment for
developmental effects
based on animal
evidence:
• Slight
Evidence from apical endpoints in in vivo mammalian animal studies considered for deriving toxicity values
Page 608 of 638
-------
Database Summary
Factors that Increase Strength
Factors that Decrease Strength
Summary of Key
Findings and within-
Stream Strength of
the Evidence
Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
• An oral gavage study evaluated
uterine parameters, number of
pups, pup weight, and viability
following gestational exposure
(GDs 7-14) in female mice
(Hazleton Laboratories. 1983).
Overall quality determination: High
Assessment of neurobehavioral and
related hormonal responses after
dosing pregnant Long-Evans rats
from GD 10 through PND 22 via
oral gavage up to 90 mg/kg-day.
No adverse effects in T3, T4, brain
or serum ACliE in dams or
offspring. No effects on brain
weight in offspring at PND 6.
Sporadic behavioral changes do not
suggest biologically relevant
adverse outcomes or developmental
toxicity. Moseret al. (2015).
Overall quality determination: High
Assessment of malformations and
variations
Kawashima et al. (1983) Overall
quality determination: Medium
Pregnant Wistar rats gavaged with
0, 50, 100, or 200 mg/kg-day from
GD 7 through 15. Twenty-tliree
percent of 30 dams at 200 mg/kg-
day died between GD 10 and 14.
At 200 mg/kg-day, male offspring
exhibited some neurotoxicity but
this was associated with significant
maternal toxicity. No other
developmental effects (e.g.,
malformations, variations, growth
effects) were observed.
• Biological gradient/dose-
response: number of litters/pair
and number of live pups/litter
decreased in a dose-related
manner during the continuous
F0 breeding phase of the
RACB.
• Supporting reproductive
effects: Magnitude and severity
of testes histological changes
increased with dose in the
subchronic dietary study in
ICR mice.
Consistency:
• Decreased numbers of live
pups/litter were observed
during continuous F0 breeding
and crossover mating phases of
the RACB.
• Decreased number of live
pups/litter was observed at the
same dose in F0 and F1
breeding phases of the RACB,
with greater severity in the
second generation.
Consistency of supporting
reproductive effects:
• Decreased testes weight was
observed in gavage and dietary
subchronic studies in mice.
• Sperm effects identified during
crossover mating correlated
with decreased fertility index
when treated males were bred
with untreated females.
• Mechanistic changes from in
vivo and in vitro studies
Consistency:
• The developmental neurotoxicity
study in rats (Moser et al„ 2015)
did not result in effects in
offspring.
The pre-natal study of
malformations and variations in
addition to neurotoxicity
(Kawashima et al„ 1983) did not
result in any significant effects in
offspring.
Note: It is possible that exposure
prior to gestation and in male
parents is needed for effects to
be observed.
Magnitude and precision:
• The developmental gavage
studies in mice used only one
dose group and no
developmental effects were
observed (Hazleton Laboratories.
1983).
Key findings'.
Available animal
toxicological studies
resulted in decreased
live pups per litter.
Overall judgment for
developmental effects
based on animal
evidence:
• Slight
Page 609 of 638
-------
Database Summary
Factors that Increase Strength
Factors that Decrease Strength
Summary of Key
Findings and within-
Stream Strength of
the Evidence
Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
(decreased testosterone,
altered steroidogenesis gene
expression) consistent with
effects on testes and sperm.
Oualitv of the database:
• Effects were observed in high-
quality studies.
Evidence in mechanistic studies and supplemental information
• Yonemoto et al. (1997) evaluated
inhibitory concentrations for cell
proliferation (IP50) and
differentiation (ID50) in rat embryo
limb bud cells.
• Reproductive: A subchronic dietaiy
study in male mice evaluated
testicular testosterone and gene
expression related to testosterone
synthesis (Chen et al.. 2015a).
• An in vitro study using TM3
Leydig cells evaluated testosterone
secretion and gene expression
related to steroidogenesis and
oxidative stress (Chen et al..
2015b).
• Three in vitro studies evaluated
estrogenic, anti-estrogenic,
androgenic, and/or anti-androgenic
activity using a yeast reporter assay
or human (endometrial, prostate
and breast) cancer cell lines
(Krivoshiev et al.. 2016; Reers et
al.. 2016; Follmann and Wober.
2006).
Biolosical eradient/dose-resoonse
(reproductive effects):
• In vivo data showed decreased
testicular testosterone and
altered gene expression related
to testosterone synthesis at the
dose in which decreased testes
weight and testicular damage
were observed.
• An in vitro study showed
decreased testosterone
secretion and/or changes in
gene expression related to
steroidogenesis and oxidative
stress at both tested
concentrations.
Consistency (Reproductive):
• Altered gene expression
related to steroidogenesis
correlated with decreased
testosterone in vivo and in
vitro.
Biolosical olausibilitv/relevance
to humans:
• Yonemoto et al. (1997)
identified an IP50 of 3600 11M
of TCEP using rat embryo
Consistency (Reproductive):
• There was inconsistency across
studies with respect to estrogen
receptor and androgen receptor
agonist and/or antagonist activity
in human (endometrial, prostate,
and breast) cancer cell lines.
Oualitv of the database:
• Few potential mechanisms were
investigated in available studies.
Biolosical olausibilitv/relevance to
humans:
• Oxidative stress is a possible
nonspecific mechanism.
Key findings: Limited
available mechanistic
data indicate that
TCEP may induce a
ratio of inhibition of
proliferation and
differentiation
resulting in concern
for development;
oxidative stress; and
endocrine disruption
via altered expression
of genes involved in
steroidogenesis.
Overall judgment for
developmental effects
based on mechanistic
evidence:
• Slight
Page 610 of 638
-------
Database Summary
Factors that Increase Strength
Factors that Decrease Strength
Summary of Key
Findings and within-
Stream Strength of
the Evidence
Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
limb bud cells. The ID50 was
1570 |iIVI: the ratio of
concentrations suggested
possible developmental
toxicity.
• Reproductive: Endocrine
disruption, via altered
expression of genes involved
in testosterone synthesis, is a
plausible mechanism for
infertility, sperm effects, and
testicular damage that is
relevant to humans.
GD = gestation day
Page 611 of 638
-------
Table Apx L-4. Evidence Integration Table for Kidney Effects
Database Summary
Factors that Increase
Strength
Factors that Decrease
Strength
Summary of Key
Findings and
within-Stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
Evidence integration summary judgment on kidney effects
Overall judgment for
renal effects based on
integration of
information across
evidence streams:
Evidence indicates that
TCEP exposure likely
causes kidney effects in
humans under relevant
exposure circumstances.
•Associations between BCEP urine levels (metabolite of
TCEP) in pregnancy and birth outcomes:
Kane et al. (2019) examined whether c\ do sure to
organophosphate esters which also included TCEP was
associated with chronic kidney disease. There was
association between BCEP levels and chronic kidney
disease related parameters (eGFR and ACR).
None
None
Key findings:
Based on one
epidemiological
study resulted in
some changes in
chronic kidney
disease related
parameters.
Overall judgment for
kidney effects based
on human evidence:
• Slight
Evidence from in vivo mammalian animal studies considered for deriving toxicity values
NTP (1991b): Rats and mice exoosed bv eavaee:
evaluated kidney weights and histopathology. Overall
quality determination: High
Kidnev weiehts (16 davs. 16 weeks, and 66 weeks 1 rats
only])
• Male rats: increased kidney weights at all time
points.
• Female rats: no change after 16 days, dose-related
increases in kidney weights after 16 weeks, and no
change after 66 weeks.
• Male mice: no change after 16 days and decreased
kidney weight after 16 weeks.
• Female mice: increased kidney weight after 16 days
and no change after 16 weeks.
Histooatholoev (16 davs. 16 weeks, and 104 weeks)
• No changes in rats or mice after 16 days or in rats
after 16 weeks.
Effect sizc/orecision:
• Histopathology
changes in rats and
mice of both sexes
were significantly
increased over controls
by both pairwise and
trend tests.
Dosc-rcsDonsc eradient:
Incidences of kidney
histopathology findings
increased with dose in rats
and mice of both sexes.
TcniDoralitv:
Histopathology findings
were more prevalent and
occurred at lower doses as
Inconsistencv
Kidney weight changes
did not occur at all time
points in female rats or
mice of either sex.
Incoherence:
Kidney weight changes
did not correspond to
histopathology changes in
female rats or mice of
either sex.
Imprecision:
• Dosing errors occurred
in 16-week studies in
rats and mice.
• Treatment-related
deaths occurred in 16-
week study in rats.
Key findings'.
Results across
available animal
toxicological studies
showed renal toxicity
in rats and mice.
Overall judgment for
renal effects based on
animal evidence:
• Moderate
Page 612 of 638
-------
Database Summary
Factors that Increase
Strength
Factors that Decrease
Strength
Summary of Key
Findings and
within-Stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
• Male rats: renal tubule hyperplasia and renal tubule
adenomas after 104 weeks at 88 mg/kg/day; one
adenoma occurred as early as 66 weeks at 88
mg/kg/day; increase in combined adenomas or
carcinomas at 88 mg/kg/day (see also TableApx L-6
for cancer endpoints).
• Female rats: renal tubule hyperplasia and renal tubule
adenomas after 104 weeks at 88 mg/kg/day (see also
Table Apx L-6 for cancer endpoints).
• Male mice: epithelial cytomegaly after 16 weeks at
700 mg/kg-day; karyomegaly after 104 weeks at
>175 mg/kg-day; one adenocarcinoma and three
adenomas at 350 mg/kg-day (see also Table Apx L-6
for cancer endpoints).
• Female mice: epithelial cytomegaly after 16 weeks at
700 mg/kg-day; karyomegaly after 104 weeks at
>175 mg/kg-day.
Taniai et al. (2012a) 28-dav eavaee studv in rats:
evaluated histopathology. Overall quality determination:
Medium
Histooatholoev
Male rats: scattered proximal tubular regeneration in the
cortex and outer stripe of the outer medulla (OSOM) at
350 mg/kg-day.
exposure duration
increased.
Consistencv:
Renal histopathology
changes were observed in
rats and mice of both
sexes and in studies in
two different laboratories.
Coherence across
endDoints:
Kidney weight changes
corresponded to
histopathology changes in
male rats.
• Survival was
decreased at the high
dose in both sexes of
rat in 104-week study.
Evidence in mechanistic studies and supplemental information
In vivo:
Markers for cell oroliferation and aoootosis (Taniai et
al.. 2012b) and reeeneratine tubules (Taniai et al..
2012a) were increased in kidnevs (OSOM and cortex)
of rats after 28 days (gavage)
In vitro:
TCEP exposure of primary rabbit renal proximal tubule
cells (PTCs) resulted in cytotoxicity, reduced DNA
Dose response sradient:
Across the in vitro
studies, dose-related
changes in the endpoints
were observed.
Consistent with related
aoical endooints: Results
from mechanistic studies
IniDrccision/Inconsistcnc
ill
• There are few studies of
mechanistic endpoints
in the kidneys.
• In vitro studies used
only one cell model and
Key findings'.
Apoptosis and altered
cell cycle regulation
may contribute to
renal effects of TCEP
in animals.
Overall judgment for
renal effects based on
Page 613 of 638
-------
Database Summary
Factors that Increase
Strength
Factors that Decrease
Strength
Summary of Key
Findings and
within-Stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
synthesis, altered expression of cell cycle regulatory
proteins, and inhibition of ion- and non-ion-transport
functions. Increased expression of pro-apoptotic
regulatory proteins and decreased expression of proteins
that inhibit aoootosis were also observed (Ren et al..
2012: Ren etal.. 2009. 2008).
are consistent with in vivo
histopathology findings in
the renal tubules.
all were conducted in
the same laboratory.
mechanistic
evidence:
• Slight
Table Apx L-5. Evidence Integration Table for Liver Effects
Database Summary
Factors that Increase Strength
Factors that Decrease Strength
Summary of Key
Findings and
Within-stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
Evidence integration summary judgment on liver effects
Evidence in studies of exposed humans considered for deriving toxicity values (none)
• Indeterminate
Overall judgment for
liver effects based on
integration of
information across
evidence streams:
Evidence suggests but is
not sufficient to
conclude that TCEP
causes hepatic effects in
humans under relevant
exposure circumstances.
Evidence from apical endpoints in in vivo mammalian animal studies for deriving toxicity values
• NTP (1991b): Subchronic and
chronic gavage studies in rats and
mice that examined liver weights,
clinical chemistry, and
histopathology. Overall quality
determination: High
• One 35-day dietary exposure study
in male mice that examined liver
weiehts (Chen et al.. 2015a).
Overall quality determination: High
Biolosical eradient/dose-
rc sdo rise:
• A dose-related trend in
hepatocellular adenoma was
observed in male mice in the
chronic study.
• Increases in liver weights in
male rats occurred at lower
doses as duration increased.
• Dose-related increases in liver
weights were seen in female
rats and female mice at 16
weeks and in male rats at 66
weeks.
Oualitv of the database:
Magnitude and precision:
• The incidence of eosinophilic
foci in male mice was
statistically significantly
increased at only the top dose
after 2 years.
Consistency:
• There were no histopathology
findings in rats or female mice,
including no hypertrophy.
• Liver weight increases were seen
in female rats after 16 days and
16 weeks, but not 66 weeks of
exposure.
• Increased liver weight was not
seen in the 35-day study.
Key findings'.
Available animal
toxicological studies
showed increased
liver weights in rats
and mice in the
absence of relevant
clinical chemistry
findings;
histopathology
changes in the liver
were observed only in
male mice.
Overall judgment for
liver effects based on
animal evidence:
Page 614 of 638
-------
Database Summary
Factors that Increase Strength
Factors that Decrease Strength
Summary of Key
Findings and
Within-stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
• Effects observed in high-
quality studies.
• No biologically relevant changes
in serum enzymes were seen in
the 2-year bioassay and not
measured in shorter studies.
Oualitv of the database:
• Liver weights were not assessed
in mice exposed longer than 16
weeks.
• Slight
Evidence in mechanistic studies and supplemental information
• One in vivo 3 5-day dietary
exposure study in male mice
examining markers of oxidative
stress (Chen et al.. 2015a).
• Five in vitro studies examining
viability, cell cycle, cellular and
mitochondrial oxidative stress,
mitochondrial function, and cell
signaling pathways in liver cells
and/or cell lines (Mennillo et al..
2019; Zhane et al.. 2017b: Zhane et
al.. 2017a: Zhane et al.. 2016c:
Zhane et al.. 2016b).
Bioloeical eradient/dose-
rc sdo rise:
• In vivo data showed induction
of hepatic oxidative stress
occurring earlier than apical
endpoints.
• Across the in vitro studies,
dose-related changes in
viability, oxidative stress, and
impaired mitochondrial
functioning were observed.
Bioloeical olausibilitv/relevance
to humans:
• Oxidative stress is a plausible
mechanism for eosinophilic
foci and tumor formation that
is relevant to humans.
Oualitv of the database:
• Few potential mechanisms were
investigated in available studies.
Bioloeical eradient/dose response:
• Oxidative stress was
demonstrated in vivo at higher
doses than those associated with
liver lesions in chronic study.
Bioloeical olausibilitv/relevance to
humans:
• Oxidative stress is a nonspecific
mechanism and was seen only at
doses higher than those
associated with liver lesions.
Key findings: Limited
available mechanistic
data indicate that
TCEP may induce
oxidative stress, alter
cellular energetics,
and/or influence cell
signaling related to
proliferation, growth,
and survival in the
liver.
Overall judgment for
liver effects based on
mechanistic evidence:
• Slight
Page 615 of 638
-------
Table Apx L-6. Evidence Integration Table for Cancer
Database Summary
Factors that Increase
Strength
Factors that Decrease
Strength
Summary of Key
Findings and
within-Stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
Evidence integration summary judgment on cancer
Evidence in studies of exposed humans considered for deriving toxicity values
Overall judgment for
cancer effects based on
integration of information
across evidence streams:
EPA concludes that
TCEP is likely to be
carcinogenic to humans
using guidance from U.S.
EPA's Guidelines for
Carcinogen Risk
Assessment (U.S. EPA.
2005b).
Hoffman et al. (2017)
Case-control study of thyroid cancer and
TCEP in household dust. Overall quality
determination: High
• Significant increase in adjusted OR (2.42)
for TCEP (in dust) above median level
among papillary thyroid cancer cases
compared to controls. TCEP in dust in
homes associated with more aggressive
tumors (n = 70 cases, 70 controls)
Liu et al. (2022) case-control studv of thvroid
cancer and TCEP in serum. Overall quality
determination: Medium
• No significant associations between TCEP
in serum and papillary thyroid cancer when
comparing quartiles of exposure.
Li et al. (2020) association between TCEP in
plasma and prevalence of gastro-intestinal and
colorectal cancers (all stages). Overall quality
determination: Medium
• TCEP was detected in GI and colorectal
cancer patients more than controls but
there were no other associations (n = 34
cases GI cancer, 40 cases colorectal cancer,
62 controls).
Liu et al. (2021) TCEP in d las ma and female-
related cancers. Overall quality determination:
Low
• No positive associations between TCEP
and cancers vs. benign tumors (n= 45
benign breast tumors, n= 73 breast cancer.
Biolosical Plausibility
• Thyroid cancers also
reported in female rats
exposed to TCEP orally.
Oualitv of the database:
• One epidemiological study
of cancer (high-quality); no
studies of renal cancers in
humans.
Biolosical eradient/dose-
rcsDonsc:
• Exposure was measured
after outcome
• TCEP concentrations
higher for benign breast
tumors compared with
breast cancer
Magnitude and Precision
• Dust used as proxy for
TCEP exposure; in on
study, corresponding
biological samples were
not collected to match with
dust samples
Consistency
• Studies using different
TCEP exposure measures
and in different geographic
regions showed different
results for papillary thyroid
cancer.
• Studied cancers not
consistently associated
with TCEP
Key findings'.
Available
epidemiological
study of cancer was
limited.
Overall judgment for
cancer effects based
on human evidence:
• Indeterminate
Page 616 of 638
-------
Database Summary
Factors that Increase
Strength
Factors that Decrease
Strength
Summary of Key
Findings and
within-Stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
n = 62 for benign uterine tumors, n = 78
for cervical cancer)
Evidence from apical endpoints in in vivo mammalian animal studies
Kidney cancer
NTP (1991b): F344 rats and B6C3F1 mice
exposed by gavage for 104 weeks. Overall
quality determination: High
• Increased incidences of adenomas and
adenomas or carcinomas in male rats (one
adenoma occurred at week 66) and
increased incidences of adenomas in
female rats.
Oualitv of the database:
• Evidence in high-quality
study in rats and mice
Magnitude and precision:
• Significant pairwise
comparisons in male and
female rats.
• Renal tubule tumors are
rare in F344/N rats and
B6C3F1 mice.
Biolosical sradient/dose-
rcsDonsc:
• Significant dose-related
trends in male and female
rats.
Consistency:
• Effects seen in both sexes
of rat.
Magnitude and precision:
• Survival was decreased at
the high dose in both sexes
of rat in 104-week study.
Consistency:
• No significantly increased
incidence of tumors was
seen in two strains of
female mice or in male
B6C3F1 mice.
Key findings'.
Dose-related
increased renal tumor
incidences
demonstrated in a
high-quality study in
rats of both sexes
Overall judgment for
kidney cancer effects
based on animal
evidence:
• Robust
Mononuclear cell leukemia
NTP (1991b): Overall aualitv determination:
High
• Increased incidence of mononuclear cell
leukemia (MNCL) in male and female rats
• No increased incidence of MNCL or other
hematologic cancer in male or female mice
Oualitv of the database:
• Evidence in high-quality
studies in rats and mice.
Magnitude and precision:
• Significant pairwise
comparisons in male and
female rats.
Biolosical sradient/dose-
rcsDonsc:
Magnitude and precision:
MNCL is common in F344
rats, its spontaneous incidence
varies widely, and incidences
in male rats exposed to TCEP
were within historical
controls.
Biolosical
plausibilitv/relevance to
humans:
Key findings'.
Dose-related
increases in MNCL
incidences
demonstrated in a
high-quality study in
rats of both sexes, but
this is a common
spontaneous cancer
in rats and only the
Page 617 of 638
-------
Database Summary
Factors that Increase
Strength
Factors that Decrease
Strength
Summary of Key
Findings and
within-Stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
• Significant dose-related
trends in male and female
rats.
Consistency:
• Evidence in two sexes.
Occurrence of MNCL is rare
in mice and other strains of
rats (Thomas et al.. 2007).
MNCL may be similar to
large granular lymphocytic
leukemia (LGLL) in humans
(Caldwell et al.. 1999;
Caldwell. 1999; Reynolds and
Foon. 1984). particularly an
aggressive form of CD3- LGL
leukemia known as aggressive
natural killer cell leukemia
(ANKCL) (Thomas et al..
2007). However. Maronoot et
al. (2016) note that ANKCL is
extremely rare with less than
98 cases reported worldwide,
and the authors contend that
ANKCL has an etiology
related to infection with
Epstein-Barr virus, not
chemical exposure.
incidence in high
dose female rats was
outside the historical
control range.
Overall judgment for
hematopoietic system
cancer effects based
on animal evidence:
• Slight
Thyroid cancer
NTP (1991b): Overall aualitv determination:
High
• Nonsignificant increase in incidence of
follicular cell adenoma or carcinoma in
male rats.
• Significantly increased incidences of
follicular cell carcinomas and adenoma or
carcinoma in female rats.
• No increased incidence of thyroid tumors
in male or female mice.
Oualitv of the database:
• Evidence in high-quality
studies in rats and mice.
Magnitude and precision:
• Significant pairwise
comparison in female rats.
Biolosical eradient/dose-
rcsDonsc:
• Significant dose-related
trend in female rats;
borderline significant trend
in males.
Magnitude and precision:
• Survival was decreased at
the high dose in both sexes
of rat in 104-week study.
Consistency:
• Effect seen in only one
species (rats).
Biolosical
plausibility/relevance to
humans:
U.S. EPA (1998a) and Dvbine
and Sanner (1999) concluded
that rodents are more sensitive
Key findings'.
Dose-related
increases in thyroid
follicular cell tumor
incidences were
demonstrated in a
high-quality study in
female rats. Rodents
may be more
sensitive than
humans to thyroid
follicular cell tumors.
Page 618 of 638
-------
Database Summary
Factors that Increase
Strength
Factors that Decrease
Strength
Summary of Key
Findings and
within-Stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
Consistency:
Effect seen in both sexes of
rats.
than humans to thyroid
follicular tumors induced by
thyroid-pituitary gland
disruption and thyroid
stimulating hormone (TSH)
hyperstimulation. NTP
(1991b) did not measure TSH
in the chronic rat study.
Overall judgment for
thyroid cancer effects
based on animal
evidence:
• Slight
Harderian gland cancer
NTP (1991b): Overall aualitv determination:
High
• Increased incidence of adenoma or
carcinoma in female mice (when interim
sacrifice groups included); no increased
incidence of Harderian gland tumors in rats
or male mice.
Oualitv of the database:
• Evidence in high-quality
studies in rats and mice.
Magnitude and precision:
• Increased incidence of
tumors in female B6C3F1
mice was statistically
significant only when
interim sacrifice groups
were included.
Biolosical eradient/dose-
rcsDonsc:
• Increased incidence in
female B6C3F1 mice
occurred only at highest
tested dose.
Consistency
• No increased incidence of
tumors in male B6C3F1
mice, or rats of either sex.
Key findings:
Increased tumor
incidence was only
seen in one sex of
one species (female
B6C3F1 mice).
Overall judgment for
Harderian gland
cancer effects based
on animal evidence:
• Slight
Page 619 of 638
-------
Database Summary
Factors that Increase
Strength
Factors that Decrease
Strength
Summary of Key
Findings and
within-Stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
Liver cancer
NTP (1991b): Overall aualitv determination:
High
• Dose-related trend for adenomas,
borderline significant increase in male
mice at high dose; no effects on female
mice or rats of either sex.
Oualitv of the database:
• Evidence in high-quality
studies in rats and mice.
Biolosical eradient/dose-
rcsDonsc:
• Significant dose-related
trend in male B6C3F1
mice.
Magnitude and precision:
• Increased incidence of
adenomas in male B6C3F1
mice was not statistically
significant by pairwise
comparison.
Consistencv
• No increase in liver tumor
incidence in female mice or
in rats of either sex.
Key findings'.
Dose-related trend in
tumor incidence was
seen only in one sex
of one species (male
B6C3F1 mice).
Overall judgment for
liver cancer effects
based on animal
evidence:
• Slight
Evidence in mechanistic studies and supplemental information
Page 620 of 638
-------
Database Summary
Factors that Increase
Strength
Factors that Decrease
Strength
Summary of Key
Findings and
within-Stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
Genotoxicity
In vivo:
• Weakly positive/equivocal for
micronucleus induction in Chinese
hamsters (Sala et al.. 1982).
In vitro:
• Positive for bacterial mutagenicity in one
S. typhimurium strains, and weakly
positive in another (Nakamura et al..
1979).
• Negative for bacterial mutagenicity in
several studies using multiple strains of S.
typhimurium with and without metabolic
activation (Follmann and Wober. 2006);
negative for mutagenicity and DNA strand
breaks in hamster V79 cells (Follmann and
Wober. 2006: Sala et al.. 1982).
• Positive for SCEs in hamster V79 cells
(Sala et al.. 1982) and DNA strand breaks
in human PBMCs (Bukowski et al.. 2019).
• Positive/weak positive for cell
transformation (may not be a genotoxic
mechanism) in two cell tvDcs (Sala et al..
1982)
Oualitv of the database:
• Tests of bacterial
mutagenicity in multiple
strains, large concentration
range, and assays with and
without metabolic
activation.
Oualitv of the database:
• Few studies in mammalian
cells and limited in vivo
data.
Magnitude and precision/
Biolosical eradient/dose-
rcsDonsc:
• Few positive findings, lack
of information on
cytotoxicity in at least one
and weak/equivocal in one.
Consistencv:
• DNA strand break findings
were not consistent across
studies/cell types.
Key findings'.
Available data
indicate that TCEP
has little genotoxic
potential. Limited
available data
indicate that TCEP
may induce oxidative
stress, alter cellular
energetics, and/or
influence cell
signaling related to
proliferation growth,
and survival in
kidney, liver, and
blood cells.
Overall judgment for
cancer effects based
on mechanistic
evidence:
• Slight
Other (non-genotoxic) mechanistic studies"
Kidnev:
• Markers for cell proliferation and apoptosis
(Taniai et al.. 2012b) and reseneratins
tubules (Taniai et al.. 2012a) were
increased in kidneys (OSOM and cortex)
of rats after 28 days (gavage)
• TCEP exposure of primary rabbit renal
proximal tubule cells (PTCs) resulted in
cytotoxicity, reduced DNA synthesis,
altered expression of cell cycle regulatory
proteins, and inhibition of ion- and non-
Biolosical eradient/dose-
rcsDonsc:
• Across the in vitro studies,
dose-related changes were
observed.
Oualitv of the database:
• There are few studies in
relevant tissue types and
only two in vivo studies.
• Available studies were not
directly focused on cancer
mechanisms.
Page 621 of 638
-------
Database Summary
Factors that Increase
Strength
Factors that Decrease
Strength
Summary of Key
Findings and
within-Stream
Strength of the
Evidence Judgment
Inferences across
Evidence Streams and
Overall Evidence
Integration Judgment
ion-transport functions. Increased
expression of pro-apoptotic regulatory
proteins and decreased expression of
proteins that inhibit apoptosis were also
observed (Ren et al.. 2012; Ren et al..
2009. 2008).
Hematopoietic:
• TCEP exposure of human peripheral blood
mononuclear cells resulted in cytotoxicity
(Mokra et al.. 2018) and decreased DNA
methvlation (Bukowski et al.. 2019).
Liver:
• Markers of oxidative stress (hepatic
antioxidant enzyme activities and their
gene expression) were increased in the
livers of male ICR mice after 35 days of
dietary c\ do sure to TCEP (Chen et al..
2015a).
• Liver cells and/or cell lines cultured with
TCEP exhibited reduced viability, cell
cycle arrest, cellular and mitochondrial
oxidative stress, impaired mitochondrial
function, and perturbation of cell signaling
oathwavs (Mennillo et al.. 2019; Zhane et
al.. 2017b; Zhane et al.. 2017a; Zhane et
al.. 2016c; Zhane et al.. 2016b).
" No tissue-specific mechanistic data related to harderian gland or thyroid follicular cell cancers were identified in the available literature.
Page 622 of 638
-------
L.2 Evidence Integration Statements for Health Outcomes with Limited
Data
Skin and Eye Irritation
The human evidence is indeterminate for skin and eye irritation. The two readily available dermal
irritation studies in animals showed inconsistent results and the single eye irritation study of medium-
quality showed that TCEP is not irritating; these studies are indeterminate. Although one study was
uninformative, EPA considered that these results are not affected by the lack of statistical analysis.
Overall, the currently available evidence is inadequate to assess whether TCEP causes irritation in
humans.
Mortality
Human evidence is indeterminate for mortality because there are no human epidemiological studies.
There is modest evidence in animal studies that shows higher mortality in rats than mice in oral studies
at the same doses and uncertain potential for mortality via the dermal route given conflicting results.
Overall, evidence suggests but is not sufficient to conclude that TCEP exposure causes mortality in
humans under relevant exposure circumstances. This conclusion is based on oral studies in rats and mice
that assessed dose levels between 12 and 700 mg/kg-day and dermal studies in rabbits at approximately
279 and 556 mg/kg-day.
Immune/Henuitological
The human evidence for immune effects based on associations between mothers' exposure during
gestation and children's respiratory symptoms is slight. Animal studies did not identify histopathological
changes in immune-related organs or in hematological parameters. A statistically significant increased
trend in mononuclear cell leukemia with increasing dose was seen in rats. In mechanistic studies, TCEP
was associated with decreases in an inflammatory cytokine and altered gene expression of inflammatory
proteins in two studies, but a third study identified inflammatory changes only after co-exposure with
benzo[a]pyrene.
Available evidence is indeterminate and therefore, is inadequate to assess whether TCEP may cause
immunological or hematological effects in humans under relevant exposure circumstances.
Thyroid
Hoffman et al. (2017) identified an association between TCEP exposure and thyroid cancer in humans
and NTP (1991b) identified increased incidences of thyroid neoplasms in rats in a 2-year cancer
bioassay but with uncertainty regarding its association with TCEP exposure. However, Moser et al.
(2015) found no changes in serum thyroid hormone levels in rat dams and offspring in a
prenatal/postnatal study. Based on these data, human evidence for thyroid effects is slight and animal
evidence is indeterminate. Overall, the currently available evidence is inadequate to assess whether
TCEP may cause thyroid changes in humans under relevant exposure circumstances.
Endocrine (Other)
Based on indeterminate human and animal evidence and lack of mechanistic support, the currently
available evidence is inadequate to assess whether TCEP may cause endocrine changes other than
thyroid and reproductive hormones in humans.
Lung/Respiratory
Two human studies that examined TCEP's association with lung function measures showed inconsistent
results, and the human evidence is indeterminate. In addition, animal data are indeterminate (no relevant
Page 623 of 638
-------
histopathological effects, lung weight changes in studies with high and uninformative overall data
quality determinations) based on high-quality studies. Therefore, the currently available evidence is
inadequate to assess whether TCEP may cause lung or respiratory effects in humans under relevant
exposure circumstances.
Body Weight
EPA identified no human studies that had information on body weight changes and therefore, human
evidence is indeterminate. In animal toxicity studies, TCEP effects on body weight were not consistent
across multiple studies. When body weight changes were observed, they were not consistently increased or
decreased. Therefore, the animal data are indeterminate. Overall, the currently available evidence is
inadequate to assess whether TCEP may cause changes in body weight in humans under relevant
exposure circumstances.
Page 624 of 638
-------
Appendix M GENOTOXICITY DATA SUMMARY
TableApx M-3 summarizes the database of studies on chromosomal aberrations, gene mutations, and
other genotoxicity endpoints for TCEP. Although EPA did not evaluate these studies using formal data
quality criteria, selected studies were reviewed by comparing against current OECD test guidelines and
important deviations are noted below. When interpreting the results of these studies, EPA also consulted
OECD (2017).
EPA did not retrieve all original studies for one or more of the following reasons: (1) they were not
readily available, (2) they were in a foreign language, (3) they evaluated effects other than chromosomal
aberrations or gene mutations, and (4) there were multiple studies of the same type (e.g., bacterial
reverse mutation assays). EPA also referred to some studies cited in the 2009 European Union Risk
Assessment Report (ECB. 2009) and Beth-Hubner (1999) for some studies that were not obtained.
M.1.1 Chromosomal Aberrations
EPA located one in vivo micronucleus assay using Chinese hamsters (Sala et al.. 1982) that was
equivocal/weakly positive for micronuclei. Two additional in vivo micronucleus studies in mice cited in
ECB (2009) and Beth-Hubner (1999) were not readily available. EPA also identified an in vitro assay
that did not find chromosomal aberrations to be associated with TCEP exposure in Chinese hamster
ovary cells (Galloway et al.. 1987).
M.l.1.1 In Vivo Data
Sala et al. (1982) report results of an in vivo micronucleus assay in which Chinese hamsters were treated
with a single i.p. dose at 0, 62.5, 125, or 250 mg/kg bw and bone marrow was evaluated for presence of
micronuclei. The authors conducted a Student's T-test to determine whether the means differed between
dose groups and the DMSO negative control. In females, the two lowest doses exhibited a statistically
significant increase in micronuclei compared with controls. Males had increased micronuclei at the
highest dose. However, only two hamsters per sex per dose were used, which would have made
statistical significance difficult to detect. When results for both sexes were combined, the two highest
doses showed differences from controls (see Table Apx M-l). The authors also conducted linear
regression to evaluate the dose response but did not report those results. The authors describe the results
as a slight effect that is difficult to interpret due to different responses between sexes and "variation with
the doses." EPA conducted a comparison of the means of each sex for each of the doses and considered
the dose-response for the combined sexes to be valid.
The study methods deviated from OECD TG 474 (OECD. 2016b) in several ways. Specifically, the
authors used an exposure route that is not recommended and scored fewer erythrocytes than
recommended (2,000 vs. 4,000). Furthermore, the study did not provide information to ensure that the
test substance reached the bone marrow, although positive effects suggest TCEP likely reached the
target tissue (Sala et al.. 1982). In addition, when using both sexes, the guidelines recommend using five
animals per sex, not two per sex. Despite these deviations, some of which might decrease the ability to
detect a response (e.g., numbers of animals/sex and number of erythrocytes scored, lack of verification
that the chemical reached the bone marrow), the results are consistent with an equivocal/ weak positive
response.
The 2009 European Union Risk Assessment Report (ECB. 2009) and Beth-Hubner (1999) reference two
additional micronucleus studies that reported negative results. The cited studies were an oral study using
NMRI mice with dosing for one time at 1,000 mg/kg and an i.p. injection study with doses up to 700
mg/kg using CD-I mice (ECB. 2009).
Page 625 of 638
-------
Table Apx M-l. Resuli
ts of In Vivo Micronucleus Test
Dose (mg/kg-bw)
Mean (Standard Deviation)*c d
Males
Females
Both Sexes
Qc7
4(1.3)
3 (0.58)
3.5 (1.0)
62.5
4 (0.82)
6.5 (1.4)*
5.25 (1.4)
125
6.25 (1.1)
7.0 (1.3)**
6.63 (1.1)***
250
7.25 (0.35)*
6.75 (3.0)
7.0(2.0)**
11 DMSO solvent control (2,200 mg/kg-bw); * p < 0.05; ** p < 0.01; *** p < 0.001
h Standard deviation is in parentheses is equal to the standard error reported in the study x square-root of n
(2/sex/dose for individual sexes and 4/dose for combined sexes)
c Number of micronuclei per 1,000 polychromatic erythrocytes
d Comparison of sexes for each does was done with the following program that compared means:
https://www.medcalc.ore/calc/comparison of means.php; the p values for 0. 62.5. 125. and 250 me/kg were
0.4252, 0.1612, 0.5969, and 0.8367, demonstrating that outcomes were not significantly different between the
sexes and the results could be combined.
Source: Salaetal. (1982)
M.l.1.2 In Vitro Data
Galloway et al. (1987) evaluated chromosomal aberrations in Chinese hamster ovary cells. Many study
methods were consistent with OECD TG 473 (OECD. 2016a). except that the authors scored only 100
cells per concentration compared with the recommended 300 per concentration needed to conclude that
a test is clearly negative. Aberrations at 0, 160, 500 and 1,600 |ig/mL were observed in 6, 10, 10 and 9
percent of cells without activation, respectively, and 4, 10, 7 and 8 percent with activation. Neither trend
test was statistically significant (p < 0.05).
M.1.2 Gene Mutations
A forward gene mutation study using Chinese hamster lung fibroblasts (Sala et al.. 1982) and multiple
bacterial reverse gene mutation assays (Follmann and Wober. 2006; Haworth et al.. 1983; BIBRA. 1977;
Prival et al.. 1977; Simmon et al.. 1977) were all negative for the induction of gene mutations. Beth-
Hubner (1999) also reported negative results in a reverse gene mutation assay yeast and in two mouse
lymphoma assays. A single study (Nakamura et al.. 1979) induced a 4- to 7-fold increase in gene
mutations in one Salmonella typhimurium strain with metabolic activation and less than a doubling in a
second strain.
M. 1.2.1 In Vitro Studies
Sala et al. (1982) evaluated the effect of TCEP exposure in a forward gene mutation assay that measured
induction of 6-thioguanine-resistant mutants using Chinese hamster lung fibroblasts (V79 cells) in the
presence and absence of metabolic activation. The authors used a negative control (acetone) as well as
two positive controls. Although the incubation times and solvents followed the recommendations of
OECD TG 476 (OECD. 2016c). the experiment did not report use of an enzyme-inducing agent for the
S9 fraction and the S9 fraction was used at 20 percent (vs. <10% as recommended by OECD TG 476).
The experiment also employed three instead of a recommended four concentrations. Furthermore, it is
not clear whether the OECD TG 476 recommended 20x 106 cells were grown by the time the cells were
treated with TCEP. The positive control run without S9 was not one of the OECD TG 476 recommended
controls. TCEP exposure did not result in increased mutations with or without S9; the authors noted that
the results were confirmed in several independent experiments.
Page 626 of 638
-------
TCEP tested negative for gene mutations in many bacterial reverse mutation assays using multiple S.
typhimurium strains (Follmann and Wober. 2006; Haworth et al.. 1983; Prival et al.. 1977; Simmon et
al.. 1977) (see TableApx M-3). Beth-Hubner (1999) references two additional studies that reported
negative results in reverse mutation assays using S. typhimurium strains TA98, TA100, TA1535,
TA1537, and TA1538.
A single study (Nakamura et al.. 1979) identified increased mutations using S. typhimurium TA100 both
with and without metabolic activation and for TA1535 in the presence of metabolic activation
(Table Apx M-3). In S. typhimurium TA100, none of the concentrations showed a doubling of
revertants compared with the negative control response. However, the TA1535 response was
approximately 4 times greater than controls at 3 |iM (-860 |ig/plate) and more than 7 times higher at 10
|iM (-2,900 |ig/plate) (Nakamura et al.. 1979). The study did not present statistical analyses. Therefore,
EPA modeled the dose-response to confirm the findings. It is not clear why the Nakamura et al. (1979)
results were inconsistent with other studies. Concentrations were comparable to other studies that
showed negative results. One difference in this study compared with others is in the method of enzyme
induction used to prepare the S9 fraction; Nakamura et al. (1979) used a mixture of PCBs (Kanechlor
500) for this induction, whereas others used Aroclor 1254 or did not appear to induce enzymes in the S9
fractions.
Table Apx M-2. Results of Bacterial Reverse Mutation Test in
Salmonella typhimurium
Concentration
(jiMol)
His+ Revertants/Plate
TA100
TA1535
-S9
+S9
-S9
+S9
0
141
140
9
14
1
158
191
14
31
3
161
192
8
57
10
172
246
6
107
30
8
86
1
7
Source: Nakamura et al. (1979)
None of the bacterial reverse mutation assays used Escherichia coli WP2 uvrA or E. coli WP2 uvrA
(PKM101), which should more likely identify oxidizing or alkylating mutagens than the Salmonella
strains used in the majority of TCEP studies. However, Follmann and Wober (2006) did test TCEP using
S. typhimurium TA102, which can also identify such mutagens, and found that TCEP did not induce
reverse mutations with this strain.
Beth-Hubner (1999) also reported negative results in a reverse gene mutation assay using
Saccharomyces cerevisiae D4 and in two mouse lymphoma assays (using the thymidine kinase locus).
M.1.3 Other Genotoxicity Assays
Table Apx M-3 summarizes two sister chromatid exchange (SCE) assays (Galloway et al.. 1987; Sala et
al.. 1982). in vitro comet assays measuring DNA damage and repair (Bukowski et al.. 2019; Follmann
and Wober. 2006). two cell transformation assays (Sala et al.. 1982). and a DNA binding assay using
TCEP (Lown et al.. 1980). Beth-Hubner (1999) also summarized an eye mosaic test (somatic mutation
and recombination) using Drosophila melanogaster.
Page 627 of 638
-------
These assays test for potentially harmful effects on genetic material such as DNA damage, cell
transformation, DNA alkylation and chromosomal damage. However, unlike gene mutation and
chromosomal aberrations studies, the changes measured in these assays may not be persistent and
transmissible.
Two studies of TCEP induction of SCEs identified equivocal results in Chinese hamster ovary cells
(positive in one of two trials with S9, negative without S9) and positive results without a dose-response
in Chinese hamster lung fibroblasts (Galloway et al.. 1987; Sala et al.. 19821 suggesting some genetic
damage, but without an understanding of the mechanism of action for this damage. The OECD test
guideline related to evaluation of SCEs (OECD TG 479) was deleted in 2014 because the mechanism for
this effect is not known (OECD. 2017).
TCEP was not considered to be an alkylating agent in an in vitro DNA binding assay (Lown et al..
1980).
Bukowski et al. (2019) conducted in vitro comet assays (alkaline and neutral) in peripheral mononuclear
blood cells (PMBCs) and identified DNA damage at the highest concentration of TCEP tested (1 mM).
Cell toxicity was not evaluated in the study, but previous results identified viability of PMBCs to be 92
percent of controls at 1 mM TCEP. DNA damage to the PMBCs was repaired within 2 hours (Bukowski
et al.. 2019). Another comet assay did not identify DNA damage in Chinese hamster fibroblasts at TCEP
concentrations up to 1 mM with or without metabolic activation (Follmann and Wober. 2006).
Sala et al. (1982) identified a high level of cell transformation in Syrian hamster embryo (SHE) cells but
a lower level with metabolic activation when using C3H10T1/2 cells. On page 24, OECD (2007) states
that "cell transformation has been related to structural alterations and changes in the expression of genes
involved in cell cycle control, proliferation and differentiation." The genomic changes may result from
direct or indirect genetic interactions or non-genotoxic mechanisms. Tamokou and Kuete (2014) notes
that the SHE assay is believed to detect early steps in the process of carcinogenesis, and that C3H10 cell
assays related to later changes.
Taniai et al. (2012a) found no statistically significant increase in immunoreactive cells associated with
repair of double-strand DNA double-strand breaks or regulation of cell cycle checkpoints after such
DNA damage in kidneys of male rats dosed with 350 mg/kg-day TCEP for 28 days.
Page 628 of 638
-------
Table Apx M-3. TCEP Genotoxicity Studies
Test Type
Exposure
Metabolic
Activation
Positive
Controls
Outcome
Reference(s)
Species (Sex)/
Route
Concentration/Dose/
Duration
Chromosomal aberrations - in vivo
Micronucleus
Chinese
hamsters
(M+F)/
intraperitoneal
0, 62.5, 125, 250
mg/kg
Single administration
NA
Yes
Equivocal, weakly
positive for micronuclei
Salaetal. (1982)
Chromosomal aberrations - in vitro
Chromosomal aberrations
Chinese
hamster ovary
cells
0, 160, 500, 1600
Hg/mL
12 lir without
activation
2 lir with activation
± S9 from rat
livers
induced with
Aroclor 1254
Yes
Negative for
chromosomal aberrations
Gallowav et al. (1987)
and NTP (1991b)
Gene mutations - in vitro
Mammalian cell forward mutation assay (6-
thioguanine-resistant mutants)
Chinese
hamster lung
fibroblasts
(V79 cells)
500, 1,000, 2,000
Hg/mL; no mention of
cytotoxicity
± S9 from rat
livers (not
induced)
Negative for mutagenicity
(both +/- S9); full results
shown only for - S9
Salaetal. (1982)
Bacterial reverse mutation assay (pre-
incubation assay)
Salmonella
tvphimurium
strains TA97a,
TA98, TA100,
TA102,
TA104,
TA1535,
TA1537,
TA1538
100 nM to 1 inM
±S9
Yes
Negative for mutagenicity
Follmann and Wober
(2006)
Bacterial reverse mutation assay (pre-
incubation assay)
Salmonella
tvphimurium
strains TA98,
TA100,
TA1535,
TA1537
0, 10,33, 100,333,
1,000, 3,333 ng/plate
± S9 from rat
and hamster
livers
induced by
Aroclor 1254
Yes,
dependent
on bacterial
strain
Not mutagenic up to toxic
doses; trials showed
toxicity /slight toxicity at
the highest dose
Haworth et al. (1983)
and
NTP (1991b)
Bacterial reverse mutation assay
Salmonella
tvphimurium
0, 1, 3, 10, 30
nM/plate
± S9 from
Kanechlor
500 (PCB)
Not
identified
Positive in TA100 and
TA1535.
Nakamura et al. (1979)
Page 629 of 638
-------
Test Type
Exposure
Metabolic
Activation
Positive
Controls
Outcome
Reference(s)
Species (Sex)/
Route
Concentration/Dose/
Duration
strains TA98,
TA100,
TA1535,
TA1537,
TA1538
[= 286.65, 859.95,
2,866.5,
8,599.5 pg/plate]
The highest concentration
showed cytotoxicity.
In vitro
bacterial reverse mutation assay
Salmonella
tvphimurium
strains TA100,
TA1535,
TA1538
1,390 and 13,900 pg/
plate "
± S9 from
normal
Sprague-
Dawley rats
and from rats
induced by
Aroclor 1254
None stated
Negative for mutagenicity
[No statistical methods
cited; visual inspection
showed lack of dose
response]
Privaletal. (1977)
In vitro
bacterial reverse mutation assay
Salmonella
tvphimurium
strains TA98,
TA100,
TA1535,
TA1537,
TA1538
Compounds were
tested up to 5
mg/plate or toxic
dose, whichever was
lower
+ S9 from
rats induced
by Aroclor
1254
[unclear
whether
TCEP was
tested
without S9]
Negative for mutagenicity
Simmon et al. (1977)
In vitro
bacterial reverse mutation assay
Salmonella
tvphimurium
strains TA 98,
TA100,
TA1535,
TA1537,
TA1538
0,0.1, 10, 100, 500,
2000 pg/plate; No
cytotoxicity observed
± S9 from
rats induced
by Aroclor
1254
Negative for mutagenicity
BIBRA (1977)
Other genotoxicity assays
In vitro Sister chromatid exchange
Cliinese
hamster ovary
cells
Without S9: One trial,
26 lir incubation
5,16,50, 160 pg/mL;
With S9. Two trials, 2
lir incubation; Trial 1:
160, 500, 1,600
pg/mL; Trial 2: 1200,
1400, 1600 pg/mL
+/-
S9 from rats
Yes
Equivocal overall
Without activation -
negative;
With activation - Trial 1
had significant responses
at the two highest doses;
Trial 2 was negative at all
doses; lowest
concentration with stat
Gallowav et al. (1987)
and NTP (1991b)
Page 630 of 638
-------
Test Type
Exposure
Metabolic
Activation
Positive
Controls
Outcome
Reference(s)
Species (Sex)/
Route
Concentration/Dose/
Duration
significant increase was
500 ug/mL; Trial 1
reached a 20% increase in
SCEs [No mention
whether cytotoxicity was
observed.]
In vitro
Sister chromatid exchanges
V79 cells
Chinese
hamster lung
fibroblasts
343, 490, 700, 1,000
Hg/ml (experiment I);
2,000, 3,000 ng/mL
(experiment II)
SCEs induced with no
clear dose response (toxic
observed at 3000 |ig/mL.
with mitosis partially
inhibited)
Salaetal. (1982)
In vitro comet assay:
DNA damage
Human:
peripheral
blood
mononuclear
cells
1 to 1,000 nM
(alkaline version)
10 to 1,000 nM
(neutral version)
Yes -
H202
(alkaline
version);
9 Gy
(neutral
version)
DNA damage observed at
1 mM in both assays
(single and double strand
breaks in alkaline version;
double strand breaks in
the neutral version).
Cell viability was not
assessed in the current
assav but Mokra et al.
(2018) identified viability
as slightly decreased at 1
mM TCEP (92% of
controls)
Bukowski et al. (2019)
In vitro comet assay:
DNA repair
Human:
peripheral
blood
mononuclear
cells
100, 500, 1,000 nM
(alkaline)
500, 1,000 nM
(neutral) for 24 hr to
induce damage; 60-
120 min for repair
assay
Single and double strand
breaks and alkali-labile
sites occurred observed at
1,000 |iIVI were repaired
after 2 hr (alkaline)
Double strand breaks at
1,000 |iIVI were repaired
after 2 hr (neutral)
Bukowski et al. (2019)
In vitro comet assay
V79 Chinese
hamster
fibroblast cells
1 to 1,000 for 24
hr
+/- S9
Yes -
potassium
dichromate
No DNA strand breaks
observed with or without
S9
Follmann and Wober
(2006)
In vitro cell transformation
Syrian hamster
embryo cells
400, 500, 600, 800
Hg/mL
High level of
transformation
Salaetal. (1982)
Page 631 of 638
-------
Test Type
Exposure
Metabolic
Activation
Positive
Controls
Outcome
Reference(s)
Species (Sex)/
Route
Concentration/Dose/
Duration
In vitro cell transformation
C3H10T1/2
cells
900 and 1,500 |ig/mL
Yes
Low incidence of
transformed foci with
metabolic activation (S9)
Salaetal. (1982)
DNA binding
In vitro
PM2-CCC-
DNA
5 inM in 180 min
No alkylation observed
Lownet al. (1980)
Page 632 of 638
-------
Appendix N
EXPOSURE RESPONSE ARRAY FOR HUMAN
HEALTH HAZARDS
The following exposure response array (Figure Apx N-l) presents HEDs for all studies and hazard
endpoints that yielded likely or suggestive evidence integration conclusions. The information is arrayed
by lowest to highest HED for NOAELs and BMDLs; all PODs based on LOAELs are listed separately.
s: UF = 30
300
rviui i c y, rtu a. m uncy w l, j.
-------
Appendix O OCCUPATIONAL EXPOSURE VALUE DERIVATION
AND ANALYTICAL METHODS USED
EPA has calculated an 8-hour time-weighted average (TWA) existing chemical occupational exposure
value to summarize the occupational exposure scenario (OES) and sensitive health endpoints into a
single value. This calculated value may be used to support risk management efforts for TCEP under
TSCA section 6(a), 15 U.S.C. §2605. EPA calculated the value rounded to 0.008 ppm (0.09 mg/m3) for
inhalation exposures to TCEP as an 8-hour TWA and for consideration in workplace settings (see
Appendix O.l) based on the lifetime cancer inhalation unit risk (IUR) for kidney cancer.
TSCA requires risk evaluations to be conducted without consideration of costs and other non-risk
factors, and thus this most sensitive occupational exposure value represents a risk-only number. If risk
management for TCEP is implemented following the risk evaluation, EPA may consider costs and other
non-risk factors, such as technological feasibility, the availability of alternatives, and the potential for
critical or essential uses. Any existing chemical exposure limit (ECEL) used for occupational safety risk
management purposes could differ from the occupational exposure value presented in this appendix
based on additional consideration of exposures and non-risk factors consistent with TSCA section 6(c).
This calculated value for TCEP represents the exposure concentration below which exposed workers
and occupational non-users are not expected to exhibit any appreciable risk of adverse toxicological
outcomes. This value accounts for potentially exposed and susceptible populations (PESS). The value is
derived based on the most sensitive human health effect (i.e., cancer) supported by the weight of
scientific evidence. This value is expressed relative to benchmarks and standard occupational scenario
assumptions of 8 hours per day, 5 days per week exposures for a total of 250 days exposure per year,
and a 40-year working life.
EPA expects that at the most sensitive occupational exposure value of 0.008 ppm (0.09 mg/m3) for
lifetime exposure, workers and occupational non-users (ONUs) also would be protected against non-
cancer effects from acute, intermediate, and chronic durations. EPA has not separately calculated a
short-term occupational exposure value (STEV) because EPA did not identify hazards for TCEP
associated with this very short exposure duration.
EPA did not identify a government-validated method for analyzing TCEP in air, but Appendix 0.2
presents a method described by La Guardia and Hale (2015) and Grimes et al. (2019). The identified limit
of detection (LOD) and limit of quantification (LOQ) using the method and the resulting monitoring
data from Grimes et al. (2019) are below the lowest calculated occupational exposure value, indicating
that monitoring below these levels may be achievable and that some workplaces may already be
achieving the occupational exposure value.
The Occupational Safety and Health Administration (OSHA) has not set a permissible exposure limit
(PEL) as an 8-hour TWA for TCEP (https://www.osha.gov/laws-
regs/regulations/standardnumber/1910/1910.1000TABLEZ2). EPA did not locate other exposure limits
for TCEP.
O.l Occupational Exposure Value Calculations
This section presents the calculations used to estimate the occupational exposure values using inputs
derived in this risk evaluation. Multiple values are presented below for hazard endpoints based on
different exposure durations (described further in section 5.2.5). For TCEP, the most sensitive
Page 634 of 638
-------
occupational exposure value is based on cancer following lifetime exposure and the resulting 8-hour
TWA is rounded to 0.008 ppm (0.09 mg/m3).
Most Sensitive Occupational Exposure Value (Lifetime Cancer)
The EVcancer is the concentration at which the extra cancer risk is equivalent to the benchmark cancer
risk of lxl0~4:
Benchmark cancer ATIUR
^resting
£ *rnnre>r * „ „ „ „ *
EVr
cancer ]UR ED * EF * WY IRworkers
lxl0-4 24^*^*78 y 0.6125^ ,
= 2 * h 25o,d * = 7.96x10 ppm
5.26x10~z per ppm 8-*^W)y 125—
d y hr
/mg\ EV ppm * MW 0.00796 ppm * 285^^ mg
cancer 3J MolarVolume 24 45 —^— 0-0928 m3
mol
Acute Non-cancer Occupational Exposure Value
The acute occupational exposure value (EVaCute) was calculated as the concentration at which the acute
margin of exposure (MOE) would equal the benchmark MOE for acute occupational exposures using the
following equation:
HECacute ATHECacute IRresting
Jh ^
Benchmark MOEacute ED IRworkers
24/l n£10rm3
4.41 ppm ~~w~~ U.olzb -7— ms
^7T— * -5TT * T- = 0.216 ppm = 2.51 —f
30 Oh m3 m3
d hr
Intermediate Non-cancer Occupational Exposure Value
The intermediate occupational exposure value (EVmtermediate) was calculated as the concentration at
which the intermediate MOE would equal the benchmark MOE for intermediate occupational exposures
using the following equation:
gy HECjntermediate ^ AThec intermediate^ ^resting
intermediate Benchmark MOfjntermediate ED*EF IRworkers
24h. m3
1 27 com ^7T* 30d 0.6125^- me
= « 4 « If = 0.0849 ppm = 0.990 ^
30 8h iOCm3 m3
-t-*22 a 1.25 -7—
a hr
The hazard value is the same for the intermediate and chronic OESs. The chronic occupational exposure
value (EVchronic) can be calculated as the concentration at which the chronic MOE would equal the
benchmark MOE for chronic occupational exposures. However, EPA has determined that because the
same key health effect applies to both intermediate and chronic exposure contexts, the relevant
averaging time should be considered equivalent across both exposure scenarios. Therefore, the resulting
EVchronic would be the same or higher than the EVmtermediate based on exposures and EPA is presenting
Only the EVmtermediate.
Page 635 of 638
-------
Where:
ATiur
ATHECacute
AT HECintermediate
AThEC chronic
Benchmarkc
= Averaging time for the cancer IUR, based on study conditions and
adjustments (24 hr/day for 365 days/yr) and averaged over a lifetime
(78 yrs) (see Risk Evaluation for Tris(2-chloroethyl) Phosphate
(TCEP) - Supplemental Information File: Supplemental Information
on Environmental Release and Occupational Exposure Assessment
(U.S. EPA 2024n) and Section 5.2.5).
= Averaging time for the POD/HEC used for evaluating non-cancer
acute occupational risk based on study conditions and HEC
adjustments (24 hr/day) (see Section 5.2.5).
= Averaging time for the POD/HEC used for evaluating non-cancer
intermediate occupational risk based on study conditions and HEC
adjustments (24 hr/day for 30 days) (see Section 5.2.5).
= Averaging time for the POD/HEC used for evaluating non-cancer
chronic occupational risk based on study conditions and HEC
adjustments (24 hr/day for 365 days/yr) (see Section 5.2.5) and
assuming the same number of years as the high-end working years
(WY, 40 years) for a worker.
= Benchmark for excess lifetime cancer risk, based on 1 x 10~4 extra risk
Benchmark MOEaCute
Benchmark MOEintermediate=
Benchmark MOEchronic =
EVc
Acute non-cancer benchmark margin of exposure, based on the total
uncertainty factor of 30 (see Section 5.2.5.1.1)
Intermediate non-cancer benchmark margin of exposure, based on the
total uncertainty factor of 30 (see Section 5.2.5.1.2)
Chronic non-cancer benchmark margin of exposure, based on the total
uncertainty factor of 30 (see Section 5.2.5.1.2)
Occupational exposure value (mg/m3 and ppm) based on lifetime
cancer risk at lxl0~4
E V acute
E V intermediate
E V chronic
ED
EF
= Occupational exposure value based on neurotoxicity from acute
exposure
= Occupational exposure value based on reproductive toxicity from
intermediate exposure
= Occupational exposure value based on reproductive toxicity from
chronic exposure
= Exposure duration (8 hr/day) (see Table 5-5)
= Exposure frequency (1 day for acute, 22 days for intermediate, and
250 days/yr for chronic and lifetime) (see Section 5.1.2.1)
Page 636 of 638
-------
HEC
Human equivalent concentration for acute, intermediate, or chronic
non-cancer occupational exposure scenarios (see Table 5-50, Table
5-51, and Table 5-52)
IUR
Inhalation unit risk (per mg/m3 and per ppm) (see Table 5-53)
IR
Inhalation rate (default is 1.25 m3/hr for workers and 0.6125 m3/hr
assumed from "resting" animals from toxicity studies)
Molar Volume
24.45 L/mol, the volume of a mole of gas at 1 atm and 25 °C
MW
Molecular weight of TCEP (285 g/mole)
WY
Working years per lifetime at the 95th percentile (40 years) (Risk
Evaluation for Tris(2-chloroethyl) Phosphate (TCEP) - Supplemental
Information File: Supplemental Information on Environmental Release
and Occupational Exposure Assessment (U.S. EPA. 2024nV)
Unit conversion:
1 ppm = 11.7 mg/m3 (see equation associated with the EVcancer calculation)
0.2.Summary of Air Sampling Analytical Methods Identified
EPA conducted a search to identify relevant NIOSH, OSHA, and EPA analytical methods used to
monitor for the presence of TCEP in air. The following sources were included for the search:
1. NIOSH Manual of Analytical Methods (NMAM); 5th Edition
URL: https://www.cdc.gov/niosh/nmam/default.html
2. NIOSH NMAM 4th Edition
URL: https://www.cdc.gov/niosh/docs/2003-154/default.html
3. OSHA Index of Sampling and Analytical Methods
URL: https://www.osha.gov/dts/sltc/methods/
4. EPA Environmental Test Method and Monitoring Information
URL: https://www.epa.gov/measurements-modeling/index-epa-test-methods
EPA did not identify any government-validated air sampling methods for TCEP. However, a method
was described and used by La Guardia and Hale (2015) and Grimes et al. (2019). The method and
associated LOD/LOQ are summarized in Table Apx 0-1.
Page 637 of 638
-------
TableApx O-l. Limit of Detection (LOD) and Limit of Quantification (LOQ) Summary for
Identified Air Sampling Analytical Methods for TCEP
Air Sampling
Analytical
Methods
Year
Published
LOD
LOQ
Notes
Source
Full-shift
personal
sampling
2019
16 ng/m3
(1.6x 1CT5
mg/m3;
1.37x 10~5
ppm)
16 ng/m3
(1.6x 1CT5
mg/m3;
1.37x 10~5
ppm)
Method reports LOD/LOQ of the
overall procedure as 16 ng/m3
(1.6x 10~5 mg/m3; 1.37x 1CT5 ppm)
using an Institute of Medicine (IOM)
sampler with a glass fiber filter at a
flow rate of 2 L/min for the inhalable
fraction of particulates and custom
OVS-2 tubes at 1 L/ per min for vapor.
Samples were sent to lab for analysis
and quantification.
Methods described
in La Guardia and
Hale (2015) and
Grimes et al.
(2019)
ppm = parts per million; ppb = parts per billion; ppt = parts per trillion
Page 638 of 638
------- |